main.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 飞桨学习赛:眼底OCT层分割-2022年12月赛段第2名\n",
    "\n",
    "---\n",
    "\n",
    "眼底OCT层分割（GOALS挑战赛任务一）\n",
    "\n",
    "## 赛题介绍\n",
    "\n",
    "光学相干断层扫描（OCT）因其无接触、非侵入性的特点，已成为眼部疾病诊疗中的常规检查，可为医生提供视网膜结构图像。与只能提供视网膜表面信息的彩色眼底图像相比，OCT图像可以提供视网膜的横断面信息，因此可以更准确地分析视网膜结构。层的分割和厚度量化对许多视网膜和视神经疾病的诊断有帮助，例如青光眼、黄斑变性或糖尿病性视网膜病变。在青光眼的诊断中，使用OCT比使用眼底彩色图像更容易发现早期病例。本学习赛是环扫OCT图像的层分割任务，目的是确定视网膜神经纤维层、神经节细胞丛层和脉络膜层，这有助于青光眼的诊断和区分。\n",
    "\n",
    "## 数据集与标签文件\n",
    "\n",
    "![](https://ai-studio-static-online.cdn.bcebos.com/e8b0f92ba6be435db7412bb04ac976860a575235ed3b40738b58012134a2051a)\n",
    "![](https://ai-studio-static-online.cdn.bcebos.com/bc95f863929840699a38c53aa4ae6b82e31bfae106ea4d25be8d9812888ab356)\n",
    "\n",
    "训练集解压文件包含如下:\n",
    "* Image 原始图片\n",
    "* Layer_Masks 标签\n",
    "* Layer_show 可视化图片（训练无需使用）\n",
    "* License-GOALS-0606.pdf 使用证书\n",
    "\n",
    "数据集是一个标准的医学影像数据集，医学影像的特点是数据少。很多时候许多图片还是来自于1个病人，并且附加的病人文本信息也作为1个输入，这样的多模态比赛在其他竞赛平台上很常见。另外1个特点是，医学影像多是灰度图。因为人体内是无法直接获取的。早期的机器学习时期就有在医学影像的应用，深度学习火热后，效果远远超越其他方法，已经被验证为一个有效的策略。\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 一.环境搭建"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-04T17:39:58.629630Z",
     "iopub.status.busy": "2023-01-04T17:39:58.628679Z",
     "iopub.status.idle": "2023-01-04T17:39:59.037128Z",
     "shell.execute_reply": "2023-01-04T17:39:59.036314Z",
     "shell.execute_reply.started": "2023-01-04T17:39:58.629592Z"
    },
    "scrolled": true,
    "tags": []
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import cv2\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-04T17:40:06.400716Z",
     "iopub.status.busy": "2023-01-04T17:40:06.399808Z",
     "iopub.status.idle": "2023-01-04T17:40:10.579850Z",
     "shell.execute_reply": "2023-01-04T17:40:10.578727Z",
     "shell.execute_reply.started": "2023-01-04T17:40:06.400680Z"
    },
    "scrolled": true,
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/assets/image_segmentation/images/\r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/assets/image_segmentation/images/human.jpg  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/assets/image_segmentation/labels/\r\n",
      " extracting: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/assets/image_segmentation/labels/label_list  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/\r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-mdpi/\r\n",
      " extracting: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-mdpi/ic_launcher.png  \r\n",
      " extracting: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-mdpi/ic_launcher_round.png  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/drawable-v24/\r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/drawable-v24/ic_launcher_foreground.xml  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-xhdpi/\r\n",
      " extracting: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-xhdpi/ic_launcher.png  \r\n",
      " extracting: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-xhdpi/ic_launcher_round.png  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-anydpi-v26/\r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-anydpi-v26/ic_launcher_round.xml  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-anydpi-v26/ic_launcher.xml  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-xxhdpi/\r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-xxhdpi/ic_launcher.png  \r\n",
      " extracting: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-xxhdpi/ic_launcher_round.png  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/drawable/\r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/drawable/ic_launcher_background.xml  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/xml/\r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/xml/settings.xml  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/values/\r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/values/styles.xml  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/values/strings.xml  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/values/arrays.xml  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/values/colors.xml  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-hdpi/\r\n",
      " extracting: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-hdpi/ic_launcher.png  \r\n",
      " extracting: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-hdpi/ic_launcher_round.png  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-xxxhdpi/\r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-xxxhdpi/ic_launcher.png  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/mipmap-xxxhdpi/ic_launcher_round.png  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/layout/\r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/layout/activity_main.xml  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/menu/\r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/res/menu/menu_action_options.xml  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/AndroidManifest.xml  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/\r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/\r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/baidu/\r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/baidu/paddle/\r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/baidu/paddle/lite/\r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/baidu/paddle/lite/demo/\r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/baidu/paddle/lite/demo/segmentation/\r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/baidu/paddle/lite/demo/segmentation/Predictor.java  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/baidu/paddle/lite/demo/segmentation/MainActivity.java  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/baidu/paddle/lite/demo/segmentation/AppCompatPreferenceActivity.java  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/baidu/paddle/lite/demo/segmentation/visual/\r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/baidu/paddle/lite/demo/segmentation/visual/Visualize.java  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/baidu/paddle/lite/demo/segmentation/SettingsActivity.java  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/baidu/paddle/lite/demo/segmentation/config/\r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/baidu/paddle/lite/demo/segmentation/config/Config.java  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/baidu/paddle/lite/demo/segmentation/preprocess/\r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/baidu/paddle/lite/demo/segmentation/preprocess/Preprocess.java  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/main/java/com/baidu/paddle/lite/demo/segmentation/Utils.java  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/androidTest/\r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/androidTest/java/\r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/androidTest/java/com/\r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/androidTest/java/com/baidu/\r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/androidTest/java/com/baidu/paddle/\r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/androidTest/java/com/baidu/paddle/lite/\r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/androidTest/java/com/baidu/paddle/lite/demo/\r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/src/androidTest/java/com/baidu/paddle/lite/demo/ExampleInstrumentedTest.java  \r\n",
      " extracting: PaddleSeg/deploy/lite/human_segmentation_demo/app/.gitignore  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/local.properties  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/build.gradle  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/app/gradlew  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/gradlew.bat  \r\n",
      " extracting: PaddleSeg/deploy/lite/human_segmentation_demo/settings.gradle  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/.gitignore  \r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/gradle/\r\n",
      "   creating: PaddleSeg/deploy/lite/human_segmentation_demo/gradle/wrapper/\r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/gradle/wrapper/gradle-wrapper.properties  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/gradle/wrapper/gradle-wrapper.jar  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/build.gradle  \r\n",
      "  inflating: PaddleSeg/deploy/lite/human_segmentation_demo/gradlew  \r\n",
      "   creating: PaddleSeg/deploy/onnxruntime_cpp/\r\n",
      "  inflating: PaddleSeg/deploy/onnxruntime_cpp/README.md  \r\n",
      "  inflating: PaddleSeg/deploy/onnxruntime_cpp/CMakeLists.txt  \r\n",
      "   creating: PaddleSeg/deploy/onnxruntime_cpp/src/\r\n",
      "  inflating: PaddleSeg/deploy/onnxruntime_cpp/src/test_seg.cpp  \r\n",
      "  inflating: PaddleSeg/deploy/onnxruntime_cpp/src/ort_session_handler.hpp  \r\n",
      "  inflating: PaddleSeg/deploy/onnxruntime_cpp/src/ort_session_handler.cpp  \r\n",
      "   creating: PaddleSeg/deploy/serving/\r\n",
      "  inflating: PaddleSeg/deploy/serving/test_serving.py  \r\n",
      "  inflating: PaddleSeg/deploy/serving/README.md  \r\n",
      "   creating: PaddleSeg/deploy/web/\r\n",
      "  inflating: PaddleSeg/deploy/web/README.md  \r\n",
      "   creating: PaddleSeg/deploy/web/example/\r\n",
      "  inflating: PaddleSeg/deploy/web/example/index.html  \r\n",
      "  inflating: PaddleSeg/deploy/web/example/index.ts  \r\n",
      "  inflating: PaddleSeg/deploy/web/example/tsconfig.json  \r\n",
      "  inflating: PaddleSeg/deploy/web/example/package.json  \r\n",
      "   creating: PaddleSeg/deploy/web/example/bg/\r\n",
      "  inflating: PaddleSeg/deploy/web/example/bg/bg.jpg  \r\n",
      "  inflating: PaddleSeg/deploy/web/example/webpack.config.js  \r\n",
      "  inflating: PaddleSeg/train.py      \r\n",
      "   creating: PaddleSeg/test_tipc/\r\n",
      "   creating: PaddleSeg/test_tipc/cpp/\r\n",
      "   creating: PaddleSeg/test_tipc/cpp/include/\r\n",
      "  inflating: PaddleSeg/test_tipc/cpp/include/preprocess_op.h  \r\n",
      "  inflating: PaddleSeg/test_tipc/cpp/include/seg.h  \r\n",
      "  inflating: PaddleSeg/test_tipc/cpp/include/config.h  \r\n",
      "  inflating: PaddleSeg/test_tipc/cpp/include/utility.h  \r\n",
      "  inflating: PaddleSeg/test_tipc/cpp/CMakeLists.txt  \r\n",
      "   creating: PaddleSeg/test_tipc/cpp/src/\r\n",
      "  inflating: PaddleSeg/test_tipc/cpp/src/preprocess_op.cpp  \r\n",
      "  inflating: PaddleSeg/test_tipc/cpp/src/main.cpp  \r\n",
      "  inflating: PaddleSeg/test_tipc/cpp/src/utility.cpp  \r\n",
      "  inflating: PaddleSeg/test_tipc/cpp/src/config.cpp  \r\n",
      "  inflating: PaddleSeg/test_tipc/cpp/src/seg.cpp  \r\n",
      "  inflating: PaddleSeg/test_tipc/cpp/cityscapes_demo.png  \r\n",
      "  inflating: PaddleSeg/test_tipc/cpp/humanseg_demo.jpg  \r\n",
      "  inflating: PaddleSeg/test_tipc/cpp/build.sh  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/\r\n",
      "   creating: PaddleSeg/test_tipc/configs/deeplabv3p_resnet50/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/deeplabv3p_resnet50/deeplabv3p_resnet50_model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/deeplabv3p_resnet50/deeplabv3p_resnet50_model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/deeplabv3p_resnet50/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/deeplabv3p_resnet50/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/deeplabv3p_resnet50/train_ptq_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/deeplabv3p_resnet50/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/deeplabv3p_resnet50/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/deeplabv3p_resnet50/deeplabv3p_resnet50_humanseg_512x512_mini_supervisely.yml  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/deeplabv3p_resnet50_cityscapes/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/deeplabv3p_resnet50_cityscapes/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/deeplabv3p_resnet50_cityscapes/deeplabv3p_resnet50_1024x512_cityscapes.yml  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/_base_/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/_base_/cityscapes_769x769_setr.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/_base_/cityscapes.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/_base_/pascal_voc12.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/_base_/cityscapes_769x769.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/_base_/ade20k.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/_base_/autonue.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/_base_/pascal_voc12aug.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/_base_/cityscapes_1024x1024.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/_base_/pascal_context.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/_base_/coco_stuff.yml  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/stdc_stdc1/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/stdc_stdc1/stdc1_seg_cityscapes_1024x512_80k.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/stdc_stdc1/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/stdc_stdc1/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/stdc_stdc1/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/stdc_stdc1/inference_cpp.txt  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc1/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc1/pp_liteseg_stdc1_cityscapes_1024x512_160k.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc1/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc1/pp_liteseg_stdc1_model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc1/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc1/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc1/train_ptq_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc1/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc1/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc1/pp_liteseg_stdc1_model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/bisenetv1/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/bisenetv1/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/bisenetv1/bisenetv1_resnet18_os8_cityscapes_1024x512_160k.yml  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/ddrnet/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ddrnet/ddrnet23_cityscapes_1024x1024_120k.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ddrnet/train_infer_python.txt  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/fcn_hrnetw18_small/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fcn_hrnetw18_small/fcn_hrnetw18_small_model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fcn_hrnetw18_small/fcn_hrnetw18_small_v1_humanseg_192x192_mini_supervisely.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fcn_hrnetw18_small/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fcn_hrnetw18_small/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fcn_hrnetw18_small/train_ptq_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fcn_hrnetw18_small/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fcn_hrnetw18_small/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fcn_hrnetw18_small/fcn_hrnetw18_small_model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/fcn_hrnetw18/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fcn_hrnetw18/fcn_hrnetw18_model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fcn_hrnetw18/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fcn_hrnetw18/fcn_hrnetw18_1024x512_cityscapes.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fcn_hrnetw18/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fcn_hrnetw18/train_ptq_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fcn_hrnetw18/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fcn_hrnetw18/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fcn_hrnetw18/fcn_hrnetw18_model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/pfpnnet/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pfpnnet/pfpn_resnet101_os8_cityscapes_512x1024_40k.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pfpnnet/train_infer_python.txt  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/encnet/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/encnet/encnet_resnet101_os8_cityscapes_1024x512_80k.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/encnet/train_infer_python.txt  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc2/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc2/pp_liteseg_stdc2_model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc2/pp_liteseg_stdc2_cityscapes_1024x512_160k.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc2/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc2/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc2/train_ptq_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc2/pp_liteseg_stdc2_model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc2/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pp_liteseg_stdc2/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/upernet/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/upernet/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/upernet/upernet_resnet101_os8_cityscapes_512x1024_40k.yml  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/espnetv2/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/espnetv2/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/espnetv2/espnet_cityscapes_1024x512_120k.yml  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/segformer_b0/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/segformer_b0/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/segformer_b0/segformer_b0_cityscapes_1024x1024_160k.yml  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/pphumanseg_lite/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pphumanseg_lite/pphumanseg_lite_model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pphumanseg_lite/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pphumanseg_lite/pphumanseg_lite_model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pphumanseg_lite/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pphumanseg_lite/train_ptq_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pphumanseg_lite/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pphumanseg_lite/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/pphumanseg_lite/pphumanseg_lite_mini_supervisely.yml  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/enet/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/enet/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/enet/enet_cityscapes_1024x512_adam_0.002_80k.yml  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/ccnet/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ccnet/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ccnet/ccnet_resnet101_os8_cityscapes_769x769_60k.yml  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/bisenetv2/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/bisenetv2/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/bisenetv2/bisenet_cityscapes_1024x1024_160k.yml  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/hrnet_w48_contrast/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/hrnet_w48_contrast/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/hrnet_w48_contrast/HRNet_W48_contrast_cityscapes_1024x512_60k.yml  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/fastscnn/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fastscnn/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/fastscnn/fastscnn_cityscapes.yml  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/ppmatting/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ppmatting/modnet_mobilenetv2.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ppmatting/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ppmatting/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ppmatting/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw48/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw48/ocrnet_hrnetw48_cityscapes_1024x512.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw48/ocrnet_hrnetw48_model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw48/ocrnet_hrnetw48_model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw48/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw48/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw48/train_ptq_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw48/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw48/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/glore/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/glore/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/glore/glore_resnet50_os8_cityscapes_1024x512_80k.yml  \r\n",
      "   creating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw18/\r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw18/ocrnet_hrnetw18_model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw18/ocrnet_hrnetw18_cityscapes_1024x512_160k.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw18/ocrnet_hrnetw18_model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw18/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw18/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw18/train_ptq_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw18/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/configs/ocrnet_hrnetw18/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt  \r\n",
      "   creating: PaddleSeg/test_tipc/serving_cpp/\r\n",
      "  inflating: PaddleSeg/test_tipc/serving_cpp/serving_client.py  \r\n",
      "  inflating: PaddleSeg/test_tipc/serving_cpp/prepare_server.sh  \r\n",
      "  inflating: PaddleSeg/test_tipc/serving_cpp/general_seg_op.h  \r\n",
      "  inflating: PaddleSeg/test_tipc/serving_cpp/modify_serving_client_conf.py  \r\n",
      "  inflating: PaddleSeg/test_tipc/serving_cpp/general_seg_op.cpp  \r\n",
      "  inflating: PaddleSeg/test_tipc/test_ptq_inference_python.sh  \r\n",
      "  inflating: PaddleSeg/test_tipc/val.py  \r\n",
      "   creating: PaddleSeg/test_tipc/results/\r\n",
      "  inflating: PaddleSeg/test_tipc/results/python_fcn_hrnetw18_small_results_fp16.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/results/python_fcn_hrnetw18_small_results_fp32.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/test_paddle2onnx.sh  \r\n",
      "  inflating: PaddleSeg/test_tipc/test_serving_infer_python.sh  \r\n",
      "  inflating: PaddleSeg/test_tipc/benchmark_train.sh  \r\n",
      "  inflating: PaddleSeg/test_tipc/compare_results.py  \r\n",
      "  inflating: PaddleSeg/test_tipc/test_serving_infer_cpp.sh  \r\n",
      "  inflating: PaddleSeg/test_tipc/requirements.txt  \r\n",
      "  inflating: PaddleSeg/test_tipc/README.md  \r\n",
      "  inflating: PaddleSeg/test_tipc/test_infer_js.sh  \r\n",
      "   creating: PaddleSeg/test_tipc/serving_python/\r\n",
      "  inflating: PaddleSeg/test_tipc/serving_python/pipeline_http_client.py  \r\n",
      "  inflating: PaddleSeg/test_tipc/serving_python/preprocess_ops.py  \r\n",
      "  inflating: PaddleSeg/test_tipc/serving_python/config.yml  \r\n",
      "  inflating: PaddleSeg/test_tipc/serving_python/web_service.py  \r\n",
      "  inflating: PaddleSeg/test_tipc/common_func.sh  \r\n",
      "   creating: PaddleSeg/test_tipc/scripts/\r\n",
      "  inflating: PaddleSeg/test_tipc/scripts/analysis.py  \r\n",
      "  inflating: PaddleSeg/test_tipc/test_inference_cpp.sh  \r\n",
      "   creating: PaddleSeg/test_tipc/web/\r\n",
      "  inflating: PaddleSeg/test_tipc/web/index.test.js  \r\n",
      "  inflating: PaddleSeg/test_tipc/web/index.html  \r\n",
      "  inflating: PaddleSeg/test_tipc/web/jest.config.js  \r\n",
      "  inflating: PaddleSeg/test_tipc/web/jest-puppeteer.config.js  \r\n",
      "   creating: PaddleSeg/test_tipc/web/imgs/\r\n",
      "  inflating: PaddleSeg/test_tipc/web/imgs/human.jpg  \r\n",
      "  inflating: PaddleSeg/test_tipc/web/imgs/seg.png  \r\n",
      "  inflating: PaddleSeg/test_tipc/prepare.sh  \r\n",
      "  inflating: PaddleSeg/test_tipc/prepare_js.sh  \r\n",
      "   creating: PaddleSeg/test_tipc/docs/\r\n",
      "  inflating: PaddleSeg/test_tipc/docs/test_train_fleet_inference_python.md  \r\n",
      "  inflating: PaddleSeg/test_tipc/docs/install.md  \r\n",
      "  inflating: PaddleSeg/test_tipc/docs/test_train_inference_python.md  \r\n",
      "  inflating: PaddleSeg/test_tipc/docs/test_infer_js.md  \r\n",
      "  inflating: PaddleSeg/test_tipc/docs/test_train_amp_inference_python.md  \r\n",
      "  inflating: PaddleSeg/test_tipc/docs/cityscapes_demo.jpg  \r\n",
      "  inflating: PaddleSeg/test_tipc/docs/guide.png  \r\n",
      "  inflating: PaddleSeg/test_tipc/docs/cityscapes_val_5.list  \r\n",
      "  inflating: PaddleSeg/test_tipc/docs/test_inference_cpp.md  \r\n",
      "  inflating: PaddleSeg/test_tipc/docs/compare_wrong.png  \r\n",
      "  inflating: PaddleSeg/test_tipc/docs/benchmark_train.md  \r\n",
      "  inflating: PaddleSeg/test_tipc/docs/test_serving_infer_python.md  \r\n",
      "  inflating: PaddleSeg/test_tipc/docs/test.png  \r\n",
      "  inflating: PaddleSeg/test_tipc/docs/test_paddle2onnx.md  \r\n",
      "  inflating: PaddleSeg/test_tipc/docs/test_serving_infer_cpp.md  \r\n",
      "  inflating: PaddleSeg/test_tipc/docs/compare_right.png  \r\n",
      "  inflating: PaddleSeg/test_tipc/test_train_inference_python.sh  \r\n",
      "  inflating: PaddleSeg/.style.yapf   \r\n",
      "   creating: PaddleSeg/docs/\r\n",
      "   creating: PaddleSeg/docs/models/\r\n",
      "  inflating: PaddleSeg/docs/models/fascnn_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/models/fascnn.md  \r\n",
      "  inflating: PaddleSeg/docs/models/ocrnet_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/models/deeplabv3_cn.md  \r\n",
      "   creating: PaddleSeg/docs/models/images/\r\n",
      "  inflating: PaddleSeg/docs/models/images/Fast-SCNN.png  \r\n",
      "  inflating: PaddleSeg/docs/models/images/convolution.png  \r\n",
      "  inflating: PaddleSeg/docs/models/images/UNet.png  \r\n",
      "  inflating: PaddleSeg/docs/models/images/deeplabv3+.png  \r\n",
      "  inflating: PaddleSeg/docs/models/images/OCRNet.png  \r\n",
      "  inflating: PaddleSeg/docs/models/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/models/unet_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/models/ocrnet.md  \r\n",
      "  inflating: PaddleSeg/docs/models/deeplabv3.md  \r\n",
      "  inflating: PaddleSeg/docs/models/unet.md  \r\n",
      "   creating: PaddleSeg/docs/deployment/\r\n",
      "   creating: PaddleSeg/docs/deployment/lite/\r\n",
      "  inflating: PaddleSeg/docs/deployment/lite/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/deployment/lite/lite_cn.md  \r\n",
      "   creating: PaddleSeg/docs/deployment/lite/example/\r\n",
      "  inflating: PaddleSeg/docs/deployment/lite/example/human_1.png  \r\n",
      "  inflating: PaddleSeg/docs/deployment/lite/example/human_2.png  \r\n",
      "  inflating: PaddleSeg/docs/deployment/lite/example/human.png  \r\n",
      "  inflating: PaddleSeg/docs/deployment/lite/example/human_3.png  \r\n",
      "  inflating: PaddleSeg/docs/deployment/lite/lite.md  \r\n",
      "   creating: PaddleSeg/docs/deployment/inference/\r\n",
      "  inflating: PaddleSeg/docs/deployment/inference/infer_benchmark_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/deployment/inference/python_inference.md  \r\n",
      "  inflating: PaddleSeg/docs/deployment/inference/inference.md  \r\n",
      "  inflating: PaddleSeg/docs/deployment/inference/inference_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/deployment/inference/infer_benchmark.md  \r\n",
      "  inflating: PaddleSeg/docs/deployment/inference/cpp_inference_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/deployment/inference/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/deployment/inference/python_inference_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/deployment/inference/cpp_inference.md  \r\n",
      "   creating: PaddleSeg/docs/deployment/serving/\r\n",
      "  inflating: PaddleSeg/docs/deployment/serving/serving.md  \r\n",
      "  inflating: PaddleSeg/docs/deployment/serving/serving_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/deployment/serving/index.rst  \r\n",
      "   creating: PaddleSeg/docs/deployment/web/\r\n",
      "  inflating: PaddleSeg/docs/deployment/web/web.md  \r\n",
      "   creating: PaddleSeg/docs/deployment/web/image/\r\n",
      "  inflating: PaddleSeg/docs/deployment/web/image/figure1.png  \r\n",
      "  inflating: PaddleSeg/docs/deployment/web/web_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/deployment/web/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/install.md  \r\n",
      "   creating: PaddleSeg/docs/train/\r\n",
      "  inflating: PaddleSeg/docs/train/train.md  \r\n",
      "  inflating: PaddleSeg/docs/train/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/train/train_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/model_export_onnx.md  \r\n",
      "  inflating: PaddleSeg/docs/loss_usage.md  \r\n",
      "   creating: PaddleSeg/docs/predict/\r\n",
      "  inflating: PaddleSeg/docs/predict/predict_cn.md  \r\n",
      "   creating: PaddleSeg/docs/predict/color_map/\r\n",
      "  inflating: PaddleSeg/docs/predict/color_map/after_mapped.jpeg  \r\n",
      "  inflating: PaddleSeg/docs/predict/color_map/before_mapped.jpeg  \r\n",
      "  inflating: PaddleSeg/docs/predict/predict.md  \r\n",
      "  inflating: PaddleSeg/docs/whole_process.md  \r\n",
      "   creating: PaddleSeg/docs/images/\r\n",
      "  inflating: PaddleSeg/docs/images/quick_start_predict.jpg  \r\n",
      "  inflating: PaddleSeg/docs/images/cityscapes_predict_demo.png  \r\n",
      "  inflating: PaddleSeg/docs/images/fig1.png  \r\n",
      "  inflating: PaddleSeg/docs/images/chat.png  \r\n",
      "  inflating: PaddleSeg/docs/images/deepglobe.png  \r\n",
      "  inflating: PaddleSeg/docs/images/love.png  \r\n",
      "   creating: PaddleSeg/docs/images/readme/\r\n",
      "  inflating: PaddleSeg/docs/images/readme/二次元.gif  \r\n",
      "  inflating: PaddleSeg/docs/images/readme/人体解析.gif  \r\n",
      "  inflating: PaddleSeg/docs/images/readme/人像分割.gif  \r\n",
      "  inflating: PaddleSeg/docs/images/readme/人像分割-0.gif  \r\n",
      "  inflating: PaddleSeg/docs/images/model.png  \r\n",
      "  inflating: PaddleSeg/docs/images/seg_news_icon.png  \r\n",
      "  inflating: PaddleSeg/docs/images/f2.png  \r\n",
      "  inflating: PaddleSeg/docs/images/anli.png  \r\n",
      "  inflating: PaddleSeg/docs/images/paddleseg_logo.png  \r\n",
      "  inflating: PaddleSeg/docs/images/fig4.png  \r\n",
      "  inflating: PaddleSeg/docs/images/f3.png  \r\n",
      "  inflating: PaddleSeg/docs/images/feature.png  \r\n",
      "  inflating: PaddleSeg/docs/images/fig3.png  \r\n",
      "  inflating: PaddleSeg/docs/images/teach.png  \r\n",
      "  inflating: PaddleSeg/docs/images/human.png  \r\n",
      "  inflating: PaddleSeg/docs/images/quick_start_vdl.jpg  \r\n",
      "  inflating: PaddleSeg/docs/images/f1.png  \r\n",
      "  inflating: PaddleSeg/docs/images/eiseg_demo.gif  \r\n",
      "  inflating: PaddleSeg/docs/images/fig2.png  \r\n",
      "  inflating: PaddleSeg/docs/images/api_fig1.png  \r\n",
      "  inflating: PaddleSeg/docs/images/interactive.gif  \r\n",
      "  inflating: PaddleSeg/docs/images/yinyong.png  \r\n",
      "  inflating: PaddleSeg/docs/images/optic_test_image.jpg  \r\n",
      "  inflating: PaddleSeg/docs/images/fig5.png  \r\n",
      "  inflating: PaddleSeg/docs/images/api_fig2.png  \r\n",
      "  inflating: PaddleSeg/docs/images/activate.png  \r\n",
      "  inflating: PaddleSeg/docs/images/Lovasz_Softmax_Evaluate_mIoU.png  \r\n",
      "  inflating: PaddleSeg/docs/images/Lovasz_Hinge_Evaluate_mIoU.png  \r\n",
      "  inflating: PaddleSeg/docs/images/QQ_chat.png  \r\n",
      "  inflating: PaddleSeg/docs/quick_start.md  \r\n",
      "  inflating: PaddleSeg/docs/model_export.md  \r\n",
      "   creating: PaddleSeg/docs/faq/\r\n",
      "   creating: PaddleSeg/docs/faq/faq/\r\n",
      "  inflating: PaddleSeg/docs/faq/faq/faq.md  \r\n",
      "  inflating: PaddleSeg/docs/faq/faq/faq_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/faq/faq/index.rst  \r\n",
      "   creating: PaddleSeg/docs/faq/faq/faq_imgs/\r\n",
      "  inflating: PaddleSeg/docs/faq/faq/faq_imgs/ann_config.png  \r\n",
      "   creating: PaddleSeg/docs/module/\r\n",
      "   creating: PaddleSeg/docs/module/images/\r\n",
      "  inflating: PaddleSeg/docs/module/images/quick_start_predict.jpg  \r\n",
      "  inflating: PaddleSeg/docs/module/images/icnet.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/cityscapes_predict_demo.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/loss_comparison.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/unet.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/poly_decay_example.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/deepglobe.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/deeplabv3p.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/data_aug_flow.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/lovasz-hinge.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/seg_news_icon.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/data_aug_example.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/VOC2012.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/pspnet.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/dice3.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/hrnet.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/piecewise_decay_example.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/gn.png  \r\n",
      " extracting: PaddleSeg/docs/module/images/qq_group2.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/usage_vis_demo.jpg  \r\n",
      "  inflating: PaddleSeg/docs/module/images/warmup_with_poly_decay_example.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/file_list2.png  \r\n",
      "   creating: PaddleSeg/docs/module/images/annotation/\r\n",
      "  inflating: PaddleSeg/docs/module/images/annotation/jingling-5.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/annotation/jingling-1.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/annotation/image-6-2.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/annotation/image-2.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/annotation/image-4-1.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/annotation/image-5.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/annotation/image-11.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/annotation/jingling-3.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/annotation/image-10.jpg  \r\n",
      "  inflating: PaddleSeg/docs/module/images/annotation/image-6.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/annotation/image-4-2.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/annotation/image-1.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/annotation/jingling-2.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/annotation/image-3.png  \r\n",
      " extracting: PaddleSeg/docs/module/images/annotation/jingling-4.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/annotation/image-7.png  \r\n",
      " extracting: PaddleSeg/docs/module/images/softmax_loss.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/file_list.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/cityscapes.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/image-10.jpg  \r\n",
      "  inflating: PaddleSeg/docs/module/images/pspnet2.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/quick_start_vdl.jpg  \r\n",
      "  inflating: PaddleSeg/docs/module/images/lovasz-softmax.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/rangescale.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/aug_method.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/fast-scnn.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/cosine_decay_example.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/data_aug_flip_mirror.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/optic_test_image.jpg  \r\n",
      "  inflating: PaddleSeg/docs/module/images/dice.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/lovasz-hinge-vis.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/visualdl_image.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/dice2.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/visualdl_scalar.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/Lovasz_Softmax_Evaluate_mIoU.png  \r\n",
      "  inflating: PaddleSeg/docs/module/images/Lovasz_Hinge_Evaluate_mIoU.png  \r\n",
      "   creating: PaddleSeg/docs/module/tricks/\r\n",
      " extracting: PaddleSeg/docs/module/tricks/tricks.md  \r\n",
      "  inflating: PaddleSeg/docs/module/tricks/index.rst  \r\n",
      "   creating: PaddleSeg/docs/module/data/\r\n",
      "  inflating: PaddleSeg/docs/module/data/data_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/data/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/module/data/data.md  \r\n",
      "   creating: PaddleSeg/docs/module/loss/\r\n",
      "  inflating: PaddleSeg/docs/module/loss/losses_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/OhemEdgeAttentionLoss_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/losses_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/OhemCrossEntropyLoss_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/lovasz_loss_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/DualTaskLoss_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/MixedLoss_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/OhemCrossEntropyLoss_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/L1Loss_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/LovaszHingeLoss_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/BCELoss_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/SemanticConnectivityLoss_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/L1Loss_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/BCELoss_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/OhemEdgeAttentionLoss_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/CrossEntropyLoss_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/EdgeAttentionLoss_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/LovaszSoftmaxLoss_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/RelaxBoundaryLoss_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/DiceLoss_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/MSELoss_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/SemanticConnectivityLoss_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/CrossEntropyLoss_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/lovasz_loss_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/BootstrappedCrossEntropyLoss_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/LovaszHingeLoss_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/DualTaskLoss_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/RelaxBoundaryLoss_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/LovaszSoftmaxLoss_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/MixedLoss_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/MSELoss_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/EdgeAttentionLoss_en.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/DiceLoss_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/loss/BootstrappedCrossEntropyLoss_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/module/index.rst  \r\n",
      "   creating: PaddleSeg/docs/pr/\r\n",
      "   creating: PaddleSeg/docs/pr/images/\r\n",
      "  inflating: PaddleSeg/docs/pr/images/001_fork.png  \r\n",
      "  inflating: PaddleSeg/docs/pr/images/002_clone.png  \r\n",
      "  inflating: PaddleSeg/docs/pr/images/004_create_pr.png  \r\n",
      "  inflating: PaddleSeg/docs/pr/images/003_precommit_pass.png  \r\n",
      "   creating: PaddleSeg/docs/pr/pr/\r\n",
      "  inflating: PaddleSeg/docs/pr/pr/pr_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/pr/pr/style_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/pr/pr/style.md  \r\n",
      "  inflating: PaddleSeg/docs/pr/pr/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/pr/pr/pr.md  \r\n",
      "  inflating: PaddleSeg/docs/release_notes.md  \r\n",
      "  inflating: PaddleSeg/docs/add_new_model.md  \r\n",
      "   creating: PaddleSeg/docs/data/\r\n",
      "   creating: PaddleSeg/docs/data/image/\r\n",
      "  inflating: PaddleSeg/docs/data/image/jingling-5.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/jingling-1.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/image-6-2.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/image-2.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/image-4-1.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/image-5.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/labelme_polygons.jpg  \r\n",
      "  inflating: PaddleSeg/docs/data/image/image-11.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/jingling-3.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/ITK-SNAP.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/file_list2.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/file_list.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/image-10.jpg  \r\n",
      "  inflating: PaddleSeg/docs/data/image/image-6.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/image-4-2.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/image-1.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/jingling-2.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/image-3.png  \r\n",
      " extracting: PaddleSeg/docs/data/image/jingling-4.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/LabelMeing.png  \r\n",
      "  inflating: PaddleSeg/docs/data/image/image-7.png  \r\n",
      "  inflating: PaddleSeg/docs/data/pre_data_cn.md  \r\n",
      "   creating: PaddleSeg/docs/data/marker/\r\n",
      "  inflating: PaddleSeg/docs/data/marker/LabelMe.md  \r\n",
      "  inflating: PaddleSeg/docs/data/marker/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/data/marker/marker_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/data/marker/LabelMe_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/data/marker/marker.md  \r\n",
      " extracting: PaddleSeg/docs/data/README.md  \r\n",
      "  inflating: PaddleSeg/docs/data/pre_data.md  \r\n",
      "   creating: PaddleSeg/docs/data/transform/\r\n",
      "  inflating: PaddleSeg/docs/data/transform/transform_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/data/transform/transform.md  \r\n",
      "  inflating: PaddleSeg/docs/data/transform/index.rst  \r\n",
      "   creating: PaddleSeg/docs/data/custom/\r\n",
      "  inflating: PaddleSeg/docs/data/custom/data_prepare_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/data/custom/data_prepare.md  \r\n",
      "  inflating: PaddleSeg/docs/data/custom/index.rst  \r\n",
      "   creating: PaddleSeg/docs/design/\r\n",
      "   creating: PaddleSeg/docs/design/use/\r\n",
      "  inflating: PaddleSeg/docs/design/use/use.md  \r\n",
      "  inflating: PaddleSeg/docs/design/use/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/design/use/use_cn.md  \r\n",
      "   creating: PaddleSeg/docs/design/create/\r\n",
      "  inflating: PaddleSeg/docs/design/create/add_new_model.md  \r\n",
      "  inflating: PaddleSeg/docs/design/create/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/design/create/add_new_model_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/requirements.txt  \r\n",
      "  inflating: PaddleSeg/docs/README.md  \r\n",
      "  inflating: PaddleSeg/docs/model_export_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/model_zoo_overview.md  \r\n",
      "   creating: PaddleSeg/docs/slim/\r\n",
      "   creating: PaddleSeg/docs/slim/distill/\r\n",
      "  inflating: PaddleSeg/docs/slim/distill/distill_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/slim/distill/distill.md  \r\n",
      "  inflating: PaddleSeg/docs/slim/distill/index.rst  \r\n",
      "   creating: PaddleSeg/docs/slim/prune/\r\n",
      "  inflating: PaddleSeg/docs/slim/prune/prune_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/slim/prune/prune.md  \r\n",
      "  inflating: PaddleSeg/docs/slim/prune/index.rst  \r\n",
      "   creating: PaddleSeg/docs/slim/quant/\r\n",
      "  inflating: PaddleSeg/docs/slim/quant/quant_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/slim/quant/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/slim/quant/quant.md  \r\n",
      "  inflating: PaddleSeg/docs/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/quick_start_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/release_notes_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/api_example.md  \r\n",
      "  inflating: PaddleSeg/docs/install_cn.md  \r\n",
      " extracting: PaddleSeg/docs/paddleseg.png  \r\n",
      "  inflating: PaddleSeg/docs/conf.py  \r\n",
      "   creating: PaddleSeg/docs/apis/\r\n",
      "  inflating: PaddleSeg/docs/apis/transforms.md  \r\n",
      "   creating: PaddleSeg/docs/apis/models/\r\n",
      "  inflating: PaddleSeg/docs/apis/models/models.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/models/models_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/models/index.rst  \r\n",
      "   creating: PaddleSeg/docs/apis/cvlibs/\r\n",
      "  inflating: PaddleSeg/docs/apis/cvlibs/cvlibs.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/cvlibs/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/apis/cvlibs/cvlibs_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/core.md  \r\n",
      "   creating: PaddleSeg/docs/apis/backbones/\r\n",
      "  inflating: PaddleSeg/docs/apis/backbones/backbones.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/backbones/backbones_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/backbones/index.rst  \r\n",
      "   creating: PaddleSeg/docs/apis/core/\r\n",
      "  inflating: PaddleSeg/docs/apis/core/core.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/core/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/apis/core/core_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/cvlibs.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/models.md  \r\n",
      "   creating: PaddleSeg/docs/apis/losses/\r\n",
      "  inflating: PaddleSeg/docs/apis/losses/losses_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/losses/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/apis/losses/losses.md  \r\n",
      "   creating: PaddleSeg/docs/apis/datasets/\r\n",
      "  inflating: PaddleSeg/docs/apis/datasets/datasets.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/datasets/datasets_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/README_CN.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/datasets.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/backbones.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/README.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/index.rst  \r\n",
      "   creating: PaddleSeg/docs/apis/transforms/\r\n",
      "  inflating: PaddleSeg/docs/apis/transforms/transforms.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/transforms/transforms_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/apis/transforms/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/apis/losses.md  \r\n",
      "  inflating: PaddleSeg/docs/api_example_cn.md  \r\n",
      "   creating: PaddleSeg/docs/evaluation/\r\n",
      "  inflating: PaddleSeg/docs/evaluation/evaluate_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/evaluation/index.rst  \r\n",
      "  inflating: PaddleSeg/docs/evaluation/evaluate.md  \r\n",
      "  inflating: PaddleSeg/docs/model_export_onnx_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/model_zoo_overview_cn.md  \r\n",
      "   creating: PaddleSeg/docs/static/\r\n",
      "  inflating: PaddleSeg/docs/static/static.md  \r\n",
      "  inflating: PaddleSeg/docs/static/static_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/make.bat  \r\n",
      "  inflating: PaddleSeg/docs/whole_process_cn.md  \r\n",
      "  inflating: PaddleSeg/docs/Makefile  \r\n",
      "   creating: PaddleSeg/contrib/\r\n",
      "   creating: PaddleSeg/contrib/DomainAdaptation/\r\n",
      "   creating: PaddleSeg/contrib/DomainAdaptation/models/\r\n",
      "   creating: PaddleSeg/contrib/DomainAdaptation/models/backbones/\r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/models/backbones/resnet.py  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/models/backbones/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/models/ema.py  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/models/gscnn.py  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/models/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/models/deeplabv2.py  \r\n",
      "   creating: PaddleSeg/contrib/DomainAdaptation/configs/\r\n",
      "   creating: PaddleSeg/contrib/DomainAdaptation/configs/deeplabv2/\r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/configs/deeplabv2/deeplabv2_resnet101_os8_gta5cityscapes_1280x640_160k_newds_edge_rec.yml  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/configs/deeplabv2/deeplabv2_resnet101_os8_gta5cityscapes_1280x640_160k_newds_edgestream.yml  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/configs/deeplabv2/deeplabv2_resnet101_os8_gta5cityscapes_1280x640_160k_newds_featpullin.yml  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/configs/deeplabv2/deeplabv2_resnet101_os8_gta5cityscapes_1280x640_160k_newds_sfnet.yml  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/configs/deeplabv2/deeplabv2_resnet101_os8_gta5cityscapes_1280x640_160k_newds_gta5src.yml  \r\n",
      "   creating: PaddleSeg/contrib/DomainAdaptation/cvlibs/\r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/cvlibs/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/cvlibs/config.py  \r\n",
      "   creating: PaddleSeg/contrib/DomainAdaptation/utils/\r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/utils/config_check.py  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/utils/utils.py  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/utils/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/utils/augmentation.py  \r\n",
      "   creating: PaddleSeg/contrib/DomainAdaptation/script/\r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/script/val.py  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/script/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/script/train.py  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/val.py  \r\n",
      "   creating: PaddleSeg/contrib/DomainAdaptation/datasets/\r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/datasets/cityscapes_noconfig.py  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/datasets/gta5_noconfig.py  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/datasets/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/datasets/synthia.py  \r\n",
      " extracting: PaddleSeg/contrib/DomainAdaptation/requirements.txt  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/README.md  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/run-DA_src.sh  \r\n",
      "  inflating: PaddleSeg/contrib/DomainAdaptation/train.py  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/\r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/configs/\r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/configs/_base_/\r\n",
      " extracting: PaddleSeg/contrib/MedicalSeg/configs/_base_/global_configs.yml  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/configs/schedulers/\r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/configs/schedulers/two_stage_coarseseg_fineseg.yml  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/configs/lung_coronavirus/\r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/configs/lung_coronavirus/lung_coronavirus.yml  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/configs/lung_coronavirus/vnet_lung_coronavirus_128_128_128_15k.yml  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/configs/lung_coronavirus/README.md  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/configs/mri_spine_seg/\r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/configs/mri_spine_seg/vnetdeepsup_mri_spine_seg_512_512_12_15k.yml  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/configs/mri_spine_seg/mri_spine_seg_1e-1_big_rmresizecrop.yml  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/configs/mri_spine_seg/README.md  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/configs/mri_spine_seg/vnet_mri_spine_seg_512_512_12_15k.yml  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/configs/mri_spine_seg/mri_spine_seg_1e-1_big_rmresizecrop_class20.yml  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/configs/msd_brain_seg/\r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/configs/msd_brain_seg/msd_brain_seg_1e-4.yml  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/configs/msd_brain_seg/README.md  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/configs/msd_brain_seg/unetr_msd_brain_seg_1e-4.yml  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/test.py  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/tools/\r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/tools/preprocess_utils/\r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/tools/preprocess_utils/values.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/tools/preprocess_utils/uncompress.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/tools/preprocess_utils/geometry.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/tools/preprocess_utils/global_var.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/tools/preprocess_utils/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/tools/preprocess_utils/load_image.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/tools/preprocess_utils/dataset_json.py  \r\n",
      " extracting: PaddleSeg/contrib/MedicalSeg/tools/preprocess_globals.yml  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/tools/prepare_prostate.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/tools/prepare_mri_spine_seg.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/tools/prepare_msd.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/tools/prepare_msd_brain_seg.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/tools/prepare.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/tools/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/tools/prepare_lung_coronavirus.py  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/medicalseg/\r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/medicalseg/models/\r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/models/unetr.py  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/medicalseg/models/losses/\r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/models/losses/cross_entropy_loss.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/models/losses/dice_loss.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/models/losses/mixes_losses.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/models/losses/loss_utils.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/models/losses/binary_cross_entropy_loss.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/models/losses/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/models/vnet_deepsup.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/models/vnet.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/models/__init__.py  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/medicalseg/cvlibs/\r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/cvlibs/manager.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/cvlibs/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/cvlibs/config.py  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/\r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/metric.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/config_check.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/train_profiler.py  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/env_util/\r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/env_util/seg_env.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/env_util/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/env_util/sys_env.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/timer.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/utils.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/progbar.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/op_flops_run.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/loss_utils.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/download.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/logger.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/utils/visualize.py  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/medicalseg/core/\r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/core/val.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/core/infer.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/core/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/core/train.py  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/medicalseg/datasets/\r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/datasets/msd_brain_seg.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/datasets/dataset.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/datasets/mri_spine_seg.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/datasets/lung_coronavirus.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/datasets/__init__.py  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/medicalseg/transforms/\r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/transforms/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/transforms/functional.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/transforms/transform.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/medicalseg/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/export.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/val.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/.pre-commit-config.yaml  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/README_CN.md  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/requirements.txt  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/README.md  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/run-vnet.sh  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/visualize.ipynb  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/LICENSE  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/.gitignore  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/documentation/\r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/documentation/tutorial_cn.md  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/documentation/tutorial.md  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/run-vnet-mri.sh  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/deploy/\r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/deploy/python/\r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/deploy/python/infer.py  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/deploy/python/README.md  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/train.py  \r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/test_tipc/\r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/test_tipc/configs/\r\n",
      "   creating: PaddleSeg/contrib/MedicalSeg/test_tipc/configs/unetr/\r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/test_tipc/configs/unetr/train_infer_python.txt  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/test_tipc/configs/unetr/msd_brain_test.yml  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/test_tipc/README.md  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/test_tipc/common_func.sh  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/test_tipc/prepare.sh  \r\n",
      "  inflating: PaddleSeg/contrib/MedicalSeg/test_tipc/test_train_inference_python.sh  \r\n",
      "   creating: PaddleSeg/contrib/CityscapesSOTA/\r\n",
      "   creating: PaddleSeg/contrib/CityscapesSOTA/models/\r\n",
      "  inflating: PaddleSeg/contrib/CityscapesSOTA/models/ocrnet_nv.py  \r\n",
      "   creating: PaddleSeg/contrib/CityscapesSOTA/models/backbones/\r\n",
      "  inflating: PaddleSeg/contrib/CityscapesSOTA/models/backbones/hrnet_nv.py  \r\n",
      "  inflating: PaddleSeg/contrib/CityscapesSOTA/models/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/CityscapesSOTA/models/mscale_ocrnet.py  \r\n",
      "   creating: PaddleSeg/contrib/CityscapesSOTA/configs/\r\n",
      "  inflating: PaddleSeg/contrib/CityscapesSOTA/configs/README.md  \r\n",
      "  inflating: PaddleSeg/contrib/CityscapesSOTA/configs/mscale_ocr_cityscapes_autolabel_mapillary_ms_val.yml  \r\n",
      "  inflating: PaddleSeg/contrib/CityscapesSOTA/configs/mscale_ocr_cityscapes_autolabel_mapillary.yml  \r\n",
      "   creating: PaddleSeg/contrib/CityscapesSOTA/tools/\r\n",
      "  inflating: PaddleSeg/contrib/CityscapesSOTA/tools/cityscapes_labels.py  \r\n",
      "  inflating: PaddleSeg/contrib/CityscapesSOTA/tools/convert_cityscapes_autolabeling.py  \r\n",
      "  inflating: PaddleSeg/contrib/CityscapesSOTA/val.py  \r\n",
      "  inflating: PaddleSeg/contrib/CityscapesSOTA/predict.py  \r\n",
      "   creating: PaddleSeg/contrib/CityscapesSOTA/datasets/\r\n",
      "  inflating: PaddleSeg/contrib/CityscapesSOTA/datasets/cityscapes_autolabeling.py  \r\n",
      "  inflating: PaddleSeg/contrib/CityscapesSOTA/datasets/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/CityscapesSOTA/README.md  \r\n",
      "   creating: PaddleSeg/contrib/CityscapesSOTA/scripts/\r\n",
      "  inflating: PaddleSeg/contrib/CityscapesSOTA/scripts/train.py  \r\n",
      "  inflating: PaddleSeg/contrib/CityscapesSOTA/train.py  \r\n",
      "   creating: PaddleSeg/contrib/AutoNUE/\r\n",
      "   creating: PaddleSeg/contrib/AutoNUE/models/\r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/models/ocrnet_nv.py  \r\n",
      "   creating: PaddleSeg/contrib/AutoNUE/models/backbones/\r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/models/backbones/hrnet_nv.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/models/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/models/mscale_ocrnet.py  \r\n",
      "   creating: PaddleSeg/contrib/AutoNUE/configs/\r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/configs/swin_transformer_mla_base_patch4_window7_160k_autonue.yml  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/configs/mscale_auto_nue_map+city@1920.yml  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/configs/auto_nue_map+city_crop.yml  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/configs/sscale_auto_nue_map+city@1920.yml  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/configs/auto_nue_auto_label.yml  \r\n",
      "   creating: PaddleSeg/contrib/AutoNUE/tools/\r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/tools/IDD_labeling.py  \r\n",
      "   creating: PaddleSeg/contrib/AutoNUE/core/\r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/core/val.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/core/infer.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/core/infer_generate_autolabel.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/core/infer_ensemble_three.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/core/predict_ensemble.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/core/infer_crop.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/core/predict_ensemble_three.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/core/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/core/val_crop.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/core/predict_generate_autolabel.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/core/infer_ensemble.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/val.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/predict.py  \r\n",
      "   creating: PaddleSeg/contrib/AutoNUE/datasets/\r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/datasets/auto_nue_crop.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/datasets/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/datasets/auto_nue_autolabel.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/datasets/auto_nue.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/README.md  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/predict_ensemble.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/predict_ensemble_three.py  \r\n",
      "   creating: PaddleSeg/contrib/AutoNUE/scripts/\r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/scripts/train.py  \r\n",
      "  inflating: PaddleSeg/contrib/AutoNUE/train.py  \r\n",
      "   creating: PaddleSeg/contrib/PanopticDeepLab/\r\n",
      "   creating: PaddleSeg/contrib/PanopticDeepLab/models/\r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/models/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/models/panoptic_deeplab.py  \r\n",
      "   creating: PaddleSeg/contrib/PanopticDeepLab/configs/\r\n",
      "   creating: PaddleSeg/contrib/PanopticDeepLab/configs/_base_/\r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/configs/_base_/cityscapes_panoptic.yml  \r\n",
      "   creating: PaddleSeg/contrib/PanopticDeepLab/configs/panoptic_deeplab/\r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/configs/panoptic_deeplab/panoptic_deeplab_resnet50_os32_cityscapes_2049x1025_bs1_90k_lr00005.yml  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/configs/panoptic_deeplab/panoptic_deeplab_resnet50_os32_cityscapes_1025x513_bs8_90k_lr00005.yml  \r\n",
      "   creating: PaddleSeg/contrib/PanopticDeepLab/utils/\r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/utils/__init__.py  \r\n",
      "   creating: PaddleSeg/contrib/PanopticDeepLab/utils/evaluation/\r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/utils/evaluation/instance.py  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/utils/evaluation/semantic.py  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/utils/evaluation/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/utils/evaluation/panoptic.py  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/utils/visualize.py  \r\n",
      "   creating: PaddleSeg/contrib/PanopticDeepLab/core/\r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/core/val.py  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/core/infer.py  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/core/predict.py  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/core/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/core/train.py  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/val.py  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/predict.py  \r\n",
      "   creating: PaddleSeg/contrib/PanopticDeepLab/datasets/\r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/datasets/cityscapes_panoptic.py  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/datasets/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/README_CN.md  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/README.md  \r\n",
      "   creating: PaddleSeg/contrib/PanopticDeepLab/transforms/\r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/transforms/target_transforms.py  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/transforms/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/train.py  \r\n",
      "   creating: PaddleSeg/contrib/PanopticDeepLab/docs/\r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/docs/visualization_semantic.png  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/docs/visualization_panoptic.png  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/docs/visualization_instance.png  \r\n",
      "  inflating: PaddleSeg/contrib/PanopticDeepLab/docs/panoptic_deeplab.jpg  \r\n",
      "   creating: PaddleSeg/contrib/PP-HumanSeg/\r\n",
      "   creating: PaddleSeg/contrib/PP-HumanSeg/configs/\r\n",
      "  inflating: PaddleSeg/contrib/PP-HumanSeg/configs/human_pp_humansegv2_lite.yml  \r\n",
      "  inflating: PaddleSeg/contrib/PP-HumanSeg/configs/portrait_pp_humansegv1_lite.yml  \r\n",
      "  inflating: PaddleSeg/contrib/PP-HumanSeg/configs/human_pp_humansegv1_server.yml  \r\n",
      "  inflating: PaddleSeg/contrib/PP-HumanSeg/configs/human_pp_humansegv1_lite.yml  \r\n",
      "  inflating: PaddleSeg/contrib/PP-HumanSeg/configs/portrait_pp_humansegv2_lite.yml  \r\n",
      "  inflating: PaddleSeg/contrib/PP-HumanSeg/configs/human_pp_humansegv2_mobile.yml  \r\n",
      "  inflating: PaddleSeg/contrib/PP-HumanSeg/configs/human_pp_humansegv1_mobile.yml  \r\n",
      "  inflating: PaddleSeg/contrib/PP-HumanSeg/README.md  \r\n",
      "  inflating: PaddleSeg/contrib/PP-HumanSeg/paper.md  \r\n",
      "   creating: PaddleSeg/contrib/PP-HumanSeg/src/\r\n",
      "  inflating: PaddleSeg/contrib/PP-HumanSeg/src/seg_demo.py  \r\n",
      "  inflating: PaddleSeg/contrib/PP-HumanSeg/src/download_pretrained_models.py  \r\n",
      "  inflating: PaddleSeg/contrib/PP-HumanSeg/src/infer.py  \r\n",
      "  inflating: PaddleSeg/contrib/PP-HumanSeg/src/download_data.py  \r\n",
      " extracting: PaddleSeg/contrib/PP-HumanSeg/src/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/PP-HumanSeg/src/optic_flow_process.py  \r\n",
      "  inflating: PaddleSeg/contrib/PP-HumanSeg/src/download_inference_models.py  \r\n",
      "  inflating: PaddleSeg/contrib/PP-HumanSeg/README_cn.md  \r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/\r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/configs/\r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/configs/fastscnn_tusimple_640x368_300k.yml  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/configs/bisenetV2_tusimple_640x368_300k.yml  \r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/core/\r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/core/val.py  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/core/infer.py  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/core/predict.py  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/core/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/core/train.py  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/export.py  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/val.py  \r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/losses/\r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/losses/lane_cross_entropy_loss.py  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/losses/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/predict.py  \r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/datasets/\r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/datasets/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/datasets/tusimple.py  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/README_CN.md  \r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/data/\r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/data/images/\r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/data/images/points/\r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/data/images/points/3.jpg  \r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/data/images/added_prediction/\r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/data/images/added_prediction/3.jpg  \r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/data/images/pseudo_color_prediction/\r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/data/images/pseudo_color_prediction/3.png  \r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/data/test_images/\r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/data/test_images/3.jpg  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/README.md  \r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/third_party/\r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/third_party/tusimple_processor.py  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/third_party/get_lane_coords.py  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/third_party/generate_tusimple_dataset.py  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/third_party/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/third_party/lane.py  \r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/transforms/\r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/transforms/__init__.py  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/transforms/lane_transforms.py  \r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/deploy/\r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/deploy/cpp/\r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/deploy/cpp/run_seg_cpu.sh  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/deploy/cpp/README_CN.md  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/deploy/cpp/README.md  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/deploy/cpp/CMakeLists.txt  \r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/deploy/cpp/src/\r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/deploy/cpp/src/lane_postprocess.hpp  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/deploy/cpp/src/lane_postprocess.cpp  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/deploy/cpp/src/test_seg.cc  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/deploy/cpp/run_seg_gpu.sh  \r\n",
      "   creating: PaddleSeg/contrib/LaneSeg/deploy/python/\r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/deploy/python/infer.py  \r\n",
      "  inflating: PaddleSeg/contrib/LaneSeg/train.py  \r\n",
      "[Errno 2] No such file or directory: 'PaddleSeg'\r\n",
      "/home/aistudio/work\r\n",
      "\u001b[31mERROR: Could not open requirements file: [Errno 2] 没有那个文件或目录: 'requirements.txt'\u001b[0m\u001b[31m\r\n",
      "\u001b[0m\r\n",
      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m A new release of pip available: \u001b[0m\u001b[31;49m22.1.2\u001b[0m\u001b[39;49m -> \u001b[0m\u001b[32;49m22.3.1\u001b[0m\r\n",
      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m To update, run: \u001b[0m\u001b[32;49mpip install --upgrade pip\u001b[0m\r\n"
     ]
    }
   ],
   "source": [
    "# 切换到工作目录\n",
    "%cd /home/aistudio/work\n",
    "# 解压下载好的PaddleSeg\n",
    "!unzip PaddleSeg.zip\n",
    "!mv PaddleSeg /home/aistudio\n",
    "# 安装PaddleSeg依赖\n",
    "%cd PaddleSeg\n",
    "!pip install -r requirements.txt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 二.数据预处理\n",
    "\n",
    "前面提到医学图像的特点，除了不同仪器产生的多模态，还有就是灰度图。对于OCT图像的像素其实还好，CT的图像分布非常大，没有足够经验的人，根本无法从原图像获取身体的各个部位。这个demo里除了PaddleSeg,主要就是额外曾加了应对这一问题的方法：windowing。同时还有噪声的问题，医学图常常伴有噪声（比如仪器造成的）。如果不做特殊处理，算法有时会把病变位置当作噪声处理掉。这也是为何使用windowing的原因。\n",
    "\n",
    "![](https://ai-studio-static-online.cdn.bcebos.com/ae41768303b242b885b55821f6e5694803d1b78eeaf04a4a9e6f39c88dc59e91)\n",
    "\n",
    "\n",
    "\n",
    "windowing是图像常用的处理手段，其作用是通过线性变换调整图像的像素值，最终使得图像易于分辨。windowing有2个参数，分别是窗宽和窗位。\n",
    "简单来说：\n",
    "\n",
    "窗宽小，对比度高，界限清晰。\n",
    "\n",
    "窗宽大，对比度低，界限模糊。\n",
    "\n",
    "窗位低，亮度高\n",
    "\n",
    "窗位高，亮度低\n",
    "\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-04T18:13:37.199113Z",
     "iopub.status.busy": "2023-01-04T18:13:37.198154Z",
     "iopub.status.idle": "2023-01-04T18:13:41.760081Z",
     "shell.execute_reply": "2023-01-04T18:13:41.758989Z",
     "shell.execute_reply.started": "2023-01-04T18:13:37.199079Z"
    },
    "scrolled": true,
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/home/aistudio/data/data170514\r\n",
      "Archive:  GOALS2022-Train.zip\r\n",
      "   creating: Train/Layer_Masks/\r\n",
      "  inflating: Train/.DS_Store         \r\n",
      "  inflating: __MACOSX/Train/._.DS_Store  \r\n",
      "   creating: Train/Image/\r\n",
      "  inflating: Train/Train_GC_GT.xlsx  \r\n",
      "  inflating: __MACOSX/Train/._Train_GC_GT.xlsx  \r\n",
      "  inflating: Train/License-GOALS-0606.pdf  \r\n",
      "  inflating: __MACOSX/Train/._License-GOALS-0606.pdf  \r\n",
      "  inflating: Train/Layer_Masks/0065.png  \r\n",
      "  inflating: Train/Layer_Masks/0071.png  \r\n",
      "  inflating: Train/Layer_Masks/0059.png  \r\n",
      "  inflating: Train/Layer_Masks/0058.png  \r\n",
      "  inflating: Train/Layer_Masks/0070.png  \r\n",
      "  inflating: Train/Layer_Masks/0064.png  \r\n",
      "  inflating: Train/Layer_Masks/0099.png  \r\n",
      "  inflating: Train/Layer_Masks/0072.png  \r\n",
      "  inflating: Train/Layer_Masks/0066.png  \r\n",
      "  inflating: Train/Layer_Masks/0067.png  \r\n",
      "  inflating: Train/Layer_Masks/0073.png  \r\n",
      "  inflating: Train/Layer_Masks/0098.png  \r\n",
      "  inflating: Train/Layer_Masks/0088.png  \r\n",
      "  inflating: Train/Layer_Masks/0077.png  \r\n",
      "  inflating: Train/Layer_Masks/0063.png  \r\n",
      "  inflating: Train/Layer_Masks/0062.png  \r\n",
      "  inflating: Train/Layer_Masks/0076.png  \r\n",
      "  inflating: Train/Layer_Masks/0089.png  \r\n",
      "  inflating: Train/Layer_Masks/0048.png  \r\n",
      "  inflating: Train/Layer_Masks/0060.png  \r\n",
      "  inflating: Train/Layer_Masks/0074.png  \r\n",
      "  inflating: Train/Layer_Masks/0100.png  \r\n",
      "  inflating: Train/Layer_Masks/0075.png  \r\n",
      "  inflating: Train/Layer_Masks/0061.png  \r\n",
      "  inflating: Train/Layer_Masks/0049.png  \r\n",
      "  inflating: Train/Layer_Masks/0006.png  \r\n",
      "  inflating: Train/Layer_Masks/0012.png  \r\n",
      "  inflating: Train/Layer_Masks/0013.png  \r\n",
      "  inflating: Train/Layer_Masks/0007.png  \r\n",
      "  inflating: Train/Layer_Masks/0011.png  \r\n",
      "  inflating: Train/Layer_Masks/0005.png  \r\n",
      "  inflating: Train/Layer_Masks/0039.png  \r\n",
      "  inflating: Train/Layer_Masks/0038.png  \r\n",
      "  inflating: Train/Layer_Masks/0004.png  \r\n",
      "  inflating: Train/Layer_Masks/0010.png  \r\n",
      "  inflating: Train/Layer_Masks/0028.png  \r\n",
      "  inflating: Train/Layer_Masks/0014.png  \r\n",
      "  inflating: Train/Layer_Masks/0001.png  \r\n",
      "  inflating: __MACOSX/Train/Layer_Masks/._0001.png  \r\n",
      "  inflating: Train/Layer_Masks/0015.png  \r\n",
      "  inflating: Train/Layer_Masks/0029.png  \r\n",
      "  inflating: Train/Layer_Masks/0003.png  \r\n",
      "  inflating: Train/Layer_Masks/0017.png  \r\n",
      "  inflating: Train/Layer_Masks/0016.png  \r\n",
      "  inflating: Train/Layer_Masks/0002.png  \r\n",
      "  inflating: Train/Layer_Masks/0027.png  \r\n",
      "  inflating: Train/Layer_Masks/0033.png  \r\n",
      "  inflating: Train/Layer_Masks/0032.png  \r\n",
      "  inflating: Train/Layer_Masks/0026.png  \r\n",
      "  inflating: Train/Layer_Masks/0030.png  \r\n",
      "  inflating: Train/Layer_Masks/0024.png  \r\n",
      "  inflating: Train/Layer_Masks/0018.png  \r\n",
      "  inflating: Train/Layer_Masks/0019.png  \r\n",
      "  inflating: Train/Layer_Masks/0025.png  \r\n",
      "  inflating: Train/Layer_Masks/0031.png  \r\n",
      "  inflating: Train/Layer_Masks/0009.png  \r\n",
      "  inflating: Train/Layer_Masks/0035.png  \r\n",
      "  inflating: Train/Layer_Masks/0021.png  \r\n",
      "  inflating: Train/Layer_Masks/0020.png  \r\n",
      "  inflating: Train/Layer_Masks/0034.png  \r\n",
      "  inflating: Train/Layer_Masks/0008.png  \r\n",
      "  inflating: Train/Layer_Masks/0022.png  \r\n",
      "  inflating: Train/Layer_Masks/0036.png  \r\n",
      "  inflating: Train/Layer_Masks/0037.png  \r\n",
      "  inflating: Train/Layer_Masks/0023.png  \r\n",
      "  inflating: Train/Layer_Masks/0087.png  \r\n",
      "  inflating: Train/Layer_Masks/0093.png  \r\n",
      "  inflating: Train/Layer_Masks/0044.png  \r\n",
      "  inflating: Train/Layer_Masks/0050.png  \r\n",
      "  inflating: Train/Layer_Masks/0078.png  \r\n",
      "  inflating: Train/Layer_Masks/0079.png  \r\n",
      "  inflating: Train/Layer_Masks/0051.png  \r\n",
      "  inflating: Train/Layer_Masks/0045.png  \r\n",
      "  inflating: Train/Layer_Masks/0092.png  \r\n",
      "  inflating: Train/Layer_Masks/0086.png  \r\n",
      "  inflating: Train/Layer_Masks/0090.png  \r\n",
      "  inflating: Train/Layer_Masks/0084.png  \r\n",
      "  inflating: Train/Layer_Masks/0053.png  \r\n",
      "  inflating: Train/Layer_Masks/0047.png  \r\n",
      "  inflating: Train/Layer_Masks/0046.png  \r\n",
      "  inflating: Train/Layer_Masks/0052.png  \r\n",
      "  inflating: Train/Layer_Masks/0085.png  \r\n",
      "  inflating: Train/Layer_Masks/0091.png  \r\n",
      "  inflating: Train/Layer_Masks/0095.png  \r\n",
      "  inflating: Train/Layer_Masks/0081.png  \r\n",
      "  inflating: Train/Layer_Masks/0056.png  \r\n",
      "  inflating: Train/Layer_Masks/0042.png  \r\n",
      "  inflating: Train/Layer_Masks/0043.png  \r\n",
      "  inflating: Train/Layer_Masks/0057.png  \r\n",
      "  inflating: Train/Layer_Masks/0080.png  \r\n",
      "  inflating: Train/Layer_Masks/0094.png  \r\n",
      "  inflating: Train/Layer_Masks/0082.png  \r\n",
      "  inflating: Train/Layer_Masks/0096.png  \r\n",
      "  inflating: Train/Layer_Masks/0069.png  \r\n",
      "  inflating: Train/Layer_Masks/0041.png  \r\n",
      "  inflating: __MACOSX/Train/Layer_Masks/._0041.png  \r\n",
      "  inflating: Train/Layer_Masks/0055.png  \r\n",
      "  inflating: Train/Layer_Masks/0054.png  \r\n",
      "  inflating: Train/Layer_Masks/0040.png  \r\n",
      "  inflating: Train/Layer_Masks/0068.png  \r\n",
      "  inflating: Train/Layer_Masks/0097.png  \r\n",
      "  inflating: Train/Layer_Masks/0083.png  \r\n",
      "  inflating: Train/Image/0065.png    \r\n",
      "  inflating: Train/Image/0071.png    \r\n",
      "  inflating: Train/Image/0059.png    \r\n",
      "  inflating: Train/Image/0058.png    \r\n",
      "  inflating: Train/Image/0070.png    \r\n",
      "  inflating: Train/Image/0064.png    \r\n",
      "  inflating: Train/Image/0099.png    \r\n",
      "  inflating: Train/Image/0072.png    \r\n",
      "  inflating: Train/Image/0066.png    \r\n",
      "  inflating: Train/Image/0067.png    \r\n",
      "  inflating: Train/Image/0073.png    \r\n",
      "  inflating: Train/Image/0098.png    \r\n",
      "  inflating: Train/Image/0088.png    \r\n",
      "  inflating: Train/Image/0077.png    \r\n",
      "  inflating: Train/Image/0063.png    \r\n",
      "  inflating: Train/Image/.DS_Store   \r\n",
      "  inflating: __MACOSX/Train/Image/._.DS_Store  \r\n",
      "  inflating: Train/Image/0062.png    \r\n",
      "  inflating: Train/Image/0076.png    \r\n",
      "  inflating: Train/Image/0089.png    \r\n",
      "  inflating: Train/Image/0048.png    \r\n",
      "  inflating: Train/Image/0060.png    \r\n",
      "  inflating: Train/Image/0074.png    \r\n",
      "  inflating: Train/Image/0100.png    \r\n",
      "  inflating: Train/Image/0075.png    \r\n",
      "  inflating: Train/Image/0061.png    \r\n",
      "  inflating: Train/Image/0049.png    \r\n",
      "  inflating: Train/Image/0006.png    \r\n",
      "  inflating: Train/Image/0012.png    \r\n",
      "  inflating: Train/Image/0013.png    \r\n",
      "  inflating: Train/Image/0007.png    \r\n",
      "  inflating: Train/Image/0011.png    \r\n",
      "  inflating: Train/Image/0005.png    \r\n",
      "  inflating: Train/Image/0039.png    \r\n",
      "  inflating: Train/Image/0038.png    \r\n",
      "  inflating: Train/Image/0004.png    \r\n",
      "  inflating: Train/Image/0010.png    \r\n",
      "  inflating: Train/Image/0028.png    \r\n",
      "  inflating: Train/Image/0014.png    \r\n",
      "  inflating: Train/Image/0001.png    \r\n",
      "  inflating: __MACOSX/Train/Image/._0001.png  \r\n",
      "  inflating: Train/Image/0015.png    \r\n",
      "  inflating: Train/Image/0029.png    \r\n",
      "  inflating: Train/Image/0003.png    \r\n",
      "  inflating: Train/Image/0017.png    \r\n",
      "  inflating: Train/Image/0016.png    \r\n",
      "  inflating: Train/Image/0002.png    \r\n",
      "  inflating: Train/Image/0027.png    \r\n",
      "  inflating: Train/Image/0033.png    \r\n",
      "  inflating: Train/Image/0032.png    \r\n",
      "  inflating: Train/Image/0026.png    \r\n",
      "  inflating: Train/Image/0030.png    \r\n",
      "  inflating: Train/Image/0024.png    \r\n",
      "  inflating: Train/Image/0018.png    \r\n",
      "  inflating: Train/Image/0019.png    \r\n",
      "  inflating: Train/Image/0025.png    \r\n",
      "  inflating: Train/Image/0031.png    \r\n",
      "  inflating: Train/Image/0009.png    \r\n",
      "  inflating: Train/Image/0035.png    \r\n",
      "  inflating: Train/Image/0021.png    \r\n",
      "  inflating: Train/Image/0020.png    \r\n",
      "  inflating: Train/Image/0034.png    \r\n",
      "  inflating: Train/Image/0008.png    \r\n",
      "  inflating: Train/Image/0022.png    \r\n",
      "  inflating: Train/Image/0036.png    \r\n",
      "  inflating: Train/Image/0037.png    \r\n",
      "  inflating: Train/Image/0023.png    \r\n",
      "  inflating: Train/Image/0087.png    \r\n",
      "  inflating: Train/Image/0093.png    \r\n",
      "  inflating: Train/Image/0044.png    \r\n",
      "  inflating: Train/Image/0050.png    \r\n",
      "  inflating: Train/Image/0078.png    \r\n",
      "  inflating: Train/Image/0079.png    \r\n",
      "  inflating: Train/Image/0051.png    \r\n",
      "  inflating: Train/Image/0045.png    \r\n",
      "  inflating: Train/Image/0092.png    \r\n",
      "  inflating: Train/Image/0086.png    \r\n",
      "  inflating: Train/Image/0090.png    \r\n",
      "  inflating: Train/Image/0084.png    \r\n",
      "  inflating: Train/Image/0053.png    \r\n",
      "  inflating: Train/Image/0047.png    \r\n",
      "  inflating: Train/Image/0046.png    \r\n",
      "  inflating: Train/Image/0052.png    \r\n",
      "  inflating: Train/Image/0085.png    \r\n",
      "  inflating: Train/Image/0091.png    \r\n",
      "  inflating: Train/Image/0095.png    \r\n",
      "  inflating: Train/Image/0081.png    \r\n",
      "  inflating: Train/Image/0056.png    \r\n",
      "  inflating: Train/Image/0042.png    \r\n",
      "  inflating: Train/Image/0043.png    \r\n",
      "  inflating: Train/Image/0057.png    \r\n",
      "  inflating: Train/Image/0080.png    \r\n",
      "  inflating: Train/Image/0094.png    \r\n",
      "  inflating: Train/Image/0082.png    \r\n",
      "  inflating: Train/Image/0096.png    \r\n",
      "  inflating: Train/Image/0069.png    \r\n",
      "  inflating: Train/Image/0041.png    \r\n",
      "  inflating: Train/Image/0055.png    \r\n",
      "  inflating: Train/Image/0054.png    \r\n",
      "  inflating: Train/Image/0040.png    \r\n",
      "  inflating: Train/Image/0068.png    \r\n",
      "  inflating: Train/Image/0097.png    \r\n",
      "  inflating: Train/Image/0083.png    \r\n",
      "  inflating: Train/Layer_show/0065.png  \r\n",
      "  inflating: Train/Layer_show/0071.png  \r\n",
      "  inflating: Train/Layer_show/0059.png  \r\n",
      "  inflating: Train/Layer_show/0058.png  \r\n",
      "  inflating: Train/Layer_show/0070.png  \r\n",
      "  inflating: Train/Layer_show/0064.png  \r\n",
      "  inflating: Train/Layer_show/0099.png  \r\n",
      "  inflating: Train/Layer_show/0072.png  \r\n",
      "  inflating: Train/Layer_show/0066.png  \r\n",
      "  inflating: Train/Layer_show/0067.png  \r\n",
      "  inflating: Train/Layer_show/0073.png  \r\n",
      "  inflating: Train/Layer_show/0098.png  \r\n",
      "  inflating: Train/Layer_show/0088.png  \r\n",
      "  inflating: Train/Layer_show/0077.png  \r\n",
      "  inflating: Train/Layer_show/0063.png  \r\n",
      "  inflating: Train/Layer_show/0062.png  \r\n",
      "  inflating: Train/Layer_show/0076.png  \r\n",
      "  inflating: Train/Layer_show/0089.png  \r\n",
      "  inflating: Train/Layer_show/0048.png  \r\n",
      "  inflating: Train/Layer_show/0060.png  \r\n",
      "  inflating: Train/Layer_show/0074.png  \r\n",
      "  inflating: Train/Layer_show/0100.png  \r\n",
      "  inflating: Train/Layer_show/0075.png  \r\n",
      "  inflating: Train/Layer_show/0061.png  \r\n",
      "  inflating: Train/Layer_show/0049.png  \r\n",
      "  inflating: Train/Layer_show/0006.png  \r\n",
      "  inflating: Train/Layer_show/0012.png  \r\n",
      "  inflating: Train/Layer_show/0013.png  \r\n",
      "  inflating: Train/Layer_show/0007.png  \r\n",
      "  inflating: Train/Layer_show/0011.png  \r\n",
      "  inflating: Train/Layer_show/0005.png  \r\n",
      "  inflating: Train/Layer_show/0039.png  \r\n",
      "  inflating: Train/Layer_show/0038.png  \r\n",
      "  inflating: Train/Layer_show/0004.png  \r\n",
      "  inflating: Train/Layer_show/0010.png  \r\n",
      "  inflating: Train/Layer_show/0028.png  \r\n",
      "  inflating: Train/Layer_show/0014.png  \r\n",
      "  inflating: Train/Layer_show/0001.png  \r\n",
      "  inflating: __MACOSX/Train/Layer_show/._0001.png  \r\n",
      "  inflating: Train/Layer_show/0015.png  \r\n",
      "  inflating: Train/Layer_show/0029.png  \r\n",
      "  inflating: Train/Layer_show/0003.png  \r\n",
      "  inflating: Train/Layer_show/0017.png  \r\n",
      "  inflating: Train/Layer_show/0016.png  \r\n",
      "  inflating: Train/Layer_show/0002.png  \r\n",
      "  inflating: Train/Layer_show/0027.png  \r\n",
      "  inflating: Train/Layer_show/0033.png  \r\n",
      "  inflating: Train/Layer_show/0032.png  \r\n",
      "  inflating: Train/Layer_show/0026.png  \r\n",
      "  inflating: Train/Layer_show/0030.png  \r\n",
      "  inflating: Train/Layer_show/0024.png  \r\n",
      "  inflating: Train/Layer_show/0018.png  \r\n",
      "  inflating: Train/Layer_show/0019.png  \r\n",
      "  inflating: Train/Layer_show/0025.png  \r\n",
      "  inflating: Train/Layer_show/0031.png  \r\n",
      "  inflating: Train/Layer_show/0009.png  \r\n",
      "  inflating: Train/Layer_show/0035.png  \r\n",
      "  inflating: Train/Layer_show/0021.png  \r\n",
      "  inflating: Train/Layer_show/0020.png  \r\n",
      "  inflating: Train/Layer_show/0034.png  \r\n",
      "  inflating: Train/Layer_show/0008.png  \r\n",
      "  inflating: Train/Layer_show/0022.png  \r\n",
      "  inflating: Train/Layer_show/0036.png  \r\n",
      "  inflating: Train/Layer_show/0037.png  \r\n",
      "  inflating: Train/Layer_show/0023.png  \r\n",
      "  inflating: Train/Layer_show/0087.png  \r\n",
      "  inflating: Train/Layer_show/0093.png  \r\n",
      "  inflating: Train/Layer_show/0044.png  \r\n",
      "  inflating: Train/Layer_show/0050.png  \r\n",
      "  inflating: Train/Layer_show/0078.png  \r\n",
      "  inflating: Train/Layer_show/0079.png  \r\n",
      "  inflating: Train/Layer_show/0051.png  \r\n",
      "  inflating: Train/Layer_show/0045.png  \r\n",
      "  inflating: Train/Layer_show/0092.png  \r\n",
      "  inflating: Train/Layer_show/0086.png  \r\n",
      "  inflating: Train/Layer_show/0090.png  \r\n",
      "  inflating: Train/Layer_show/0084.png  \r\n",
      "  inflating: Train/Layer_show/0053.png  \r\n",
      "  inflating: Train/Layer_show/0047.png  \r\n",
      "  inflating: Train/Layer_show/0046.png  \r\n",
      "  inflating: Train/Layer_show/0052.png  \r\n",
      "  inflating: Train/Layer_show/0085.png  \r\n",
      "  inflating: Train/Layer_show/0091.png  \r\n",
      "  inflating: Train/Layer_show/0095.png  \r\n",
      "  inflating: Train/Layer_show/0081.png  \r\n",
      "  inflating: Train/Layer_show/0056.png  \r\n",
      "  inflating: Train/Layer_show/0042.png  \r\n",
      "  inflating: Train/Layer_show/0043.png  \r\n",
      "  inflating: Train/Layer_show/0057.png  \r\n",
      "  inflating: Train/Layer_show/0080.png  \r\n",
      "  inflating: Train/Layer_show/0094.png  \r\n",
      "  inflating: Train/Layer_show/0082.png  \r\n",
      "  inflating: Train/Layer_show/0096.png  \r\n",
      "  inflating: Train/Layer_show/0069.png  \r\n",
      "  inflating: Train/Layer_show/0041.png  \r\n",
      "  inflating: __MACOSX/Train/Layer_show/._0041.png  \r\n",
      "  inflating: Train/Layer_show/0055.png  \r\n",
      "  inflating: Train/Layer_show/0054.png  \r\n",
      "  inflating: Train/Layer_show/0040.png  \r\n",
      "  inflating: Train/Layer_show/0068.png  \r\n",
      "  inflating: Train/Layer_show/0097.png  \r\n",
      "  inflating: Train/Layer_show/0083.png  \r\n",
      "Archive:  GOALS2022-Validation.zip\r\n",
      "  inflating: GOALS2022-Validation/.DS_Store  \r\n",
      "  inflating: __MACOSX/GOALS2022-Validation/._.DS_Store  \r\n",
      "   creating: GOALS2022-Validation/Image/\r\n",
      "  inflating: GOALS2022-Validation/License-GOALS-0606.pdf  \r\n",
      "  inflating: __MACOSX/GOALS2022-Validation/._License-GOALS-0606.pdf  \r\n",
      "  inflating: GOALS2022-Validation/Image/0111.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0105.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0139.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0138.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0104.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0110.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0106.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0112.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0113.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0107.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0103.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0117.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/.DS_Store  \r\n",
      "  inflating: __MACOSX/GOALS2022-Validation/Image/._.DS_Store  \r\n",
      "  inflating: GOALS2022-Validation/Image/0116.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0102.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0128.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0114.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0101.png  \r\n",
      "  inflating: __MACOSX/GOALS2022-Validation/Image/._0101.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0115.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0129.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0199.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0172.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0166.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0167.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0173.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0198.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0165.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0171.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0159.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0158.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0170.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0164.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0148.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0160.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0174.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0175.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0161.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0149.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0188.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0177.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0163.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0162.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0176.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0189.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0200.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0190.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0184.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0153.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0147.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0146.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0152.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0185.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0191.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0187.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0193.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0144.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0150.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0178.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0179.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0151.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0145.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0192.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0186.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0182.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0196.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0169.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0141.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0155.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0154.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0140.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0168.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0197.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0183.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0195.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0181.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0156.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0142.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0143.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0157.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0180.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0194.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0130.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0124.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0118.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0119.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0125.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0131.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0127.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0133.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0132.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0126.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0122.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0136.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0137.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0123.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0109.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0135.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0121.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0120.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0134.png  \r\n",
      "  inflating: GOALS2022-Validation/Image/0108.png  \r\n"
     ]
    }
   ],
   "source": [
    "# 解压数据集\n",
    "%cd /home/aistudio/data/data170514\n",
    "\n",
    "!unzip -o GOALS2022-Train.zip\n",
    "!unzip -o GOALS2022-Validation.zip\n",
    "\n",
    "! mv Train/Image /home/aistudio/PaddleSeg/data/mine/\n",
    "! mv Train/Layer_Masks /home/aistudio/PaddleSeg/data/mine/\n",
    "! mv GOALS2022-Validation/Image /home/aistudio/PaddleSeg/data/mine/Test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "统计所有图像的像素值分布，进而获取原始数据集中的窗宽分布。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-04T18:13:48.774130Z",
     "iopub.status.busy": "2023-01-04T18:13:48.773365Z",
     "iopub.status.idle": "2023-01-04T18:13:53.835515Z",
     "shell.execute_reply": "2023-01-04T18:13:53.834560Z",
     "shell.execute_reply.started": "2023-01-04T18:13:48.774092Z"
    },
    "scrolled": true,
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(800, 110000)\r\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "datapath='/home/aistudio/PaddleSeg/data/mine/' \n",
    "data=cv2.imread(datapath+'Image/0001.png',0)\n",
    "for i in range(2,101):\n",
    "        name = str(i).zfill(4)\n",
    "        image = cv2.imread(datapath+'Image/'+name+'.png',0)\n",
    "        data=np.concatenate((data,image),axis=1)\n",
    "print(data.shape)\n",
    "plt.hist(data.reshape(-1,),bins=50)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-04T18:14:01.795648Z",
     "iopub.status.busy": "2023-01-04T18:14:01.794515Z",
     "iopub.status.idle": "2023-01-04T18:14:01.800321Z",
     "shell.execute_reply": "2023-01-04T18:14:01.799611Z",
     "shell.execute_reply.started": "2023-01-04T18:14:01.795612Z"
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "#给定windowing自定义函数\n",
    "def windowing(img, window_width, window_center):\n",
    "    #img： 需要增强的图片\n",
    "    #window_width:窗宽\n",
    "    #window_center:中心\n",
    "    minWindow = float(window_center)-0.5*float(window_width)\n",
    "    new_img = (img-minWindow)/float(window_width)\n",
    "    new_img[new_img<0] = 0\n",
    "    new_img[new_img>1] = 1\n",
    "    return (new_img*255).astype('uint8') #把数据整理成标准图像格式"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-04T18:14:04.568560Z",
     "iopub.status.busy": "2023-01-04T18:14:04.568032Z",
     "iopub.status.idle": "2023-01-04T18:14:04.952358Z",
     "shell.execute_reply": "2023-01-04T18:14:04.951612Z",
     "shell.execute_reply.started": "2023-01-04T18:14:04.568527Z"
    },
    "scrolled": true,
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data1=cv2.imread(datapath+'Image/0001.png',0)\n",
    "img_ct = windowing(data1,200,100)\n",
    "plt.imshow(img_ct,cmap='gray')\n",
    "plt.show()\n",
    "cv2.imwrite(\"/home/aistudio/test.png\",img_ct)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-04T18:14:08.708587Z",
     "iopub.status.busy": "2023-01-04T18:14:08.707687Z",
     "iopub.status.idle": "2023-01-04T18:14:09.256963Z",
     "shell.execute_reply": "2023-01-04T18:14:09.256221Z",
     "shell.execute_reply.started": "2023-01-04T18:14:08.708552Z"
    },
    "scrolled": true,
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 9 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#依次增大图像的位宽，图像对比度降低。\n",
    "#如果窗宽过低，部分低的像素会直接区分不出背景。所以对比度不能过高，也不能过低。\n",
    "j = 1\n",
    "data1=cv2.imread(datapath+'Image/0001.png',0)\n",
    "for i in range(1,10):\n",
    "    plt.subplot(3,3,j)\n",
    "    img_ct = windowing(data1,i*30,150)\n",
    "    plt.imshow(img_ct,cmap='gray')\n",
    "    plt.axis('off')\n",
    "    j += 1\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-04T18:14:12.219834Z",
     "iopub.status.busy": "2023-01-04T18:14:12.218920Z",
     "iopub.status.idle": "2023-01-04T18:14:12.832579Z",
     "shell.execute_reply": "2023-01-04T18:14:12.831729Z",
     "shell.execute_reply.started": "2023-01-04T18:14:12.219801Z"
    },
    "scrolled": true,
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 9 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#依次增大图像的窗中心（窗位的另外一个叫法），亮度降低\n",
    "j = 1\n",
    "data1=cv2.imread(datapath+'Image/0001.png',0)\n",
    "for i in range(1,10):\n",
    "    plt.subplot(3,3,j)\n",
    "    img_ct = windowing(data1,150,i*30)\n",
    "    plt.imshow(img_ct,cmap='gray')\n",
    "    plt.axis('off')\n",
    "    j += 1\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-04T18:14:16.894738Z",
     "iopub.status.busy": "2023-01-04T18:14:16.894197Z",
     "iopub.status.idle": "2023-01-04T18:14:20.243543Z",
     "shell.execute_reply": "2023-01-04T18:14:20.242445Z",
     "shell.execute_reply.started": "2023-01-04T18:14:16.894705Z"
    },
    "scrolled": true,
    "tags": []
   },
   "outputs": [],
   "source": [
    "!mkdir /home/aistudio/PaddleSeg/data/mine/image\n",
    "class windowImg(object):\n",
    "    def __init__(self,mode='train'):\n",
    "        self.dataPath = '/home/aistudio/PaddleSeg/data/mine/'\n",
    "        self.mode=mode\n",
    "    def run(self):\n",
    "        if self.mode=='train':\n",
    "            for i in range(1,101):\n",
    "                name = str(i).zfill(4)\n",
    "                image = cv2.imread(self.dataPath+'Image/'+name+'.png',0)\n",
    "                image = windowing(image,200,100)\n",
    "                cv2.imwrite(self.dataPath+'image/'+name+'.png', image)\n",
    "        elif self.mode=='test':\n",
    "            for i in range(101,201):\n",
    "                name = str(i).zfill(4)\n",
    "                image = cv2.imread(self.dataPath+'Test/'+name+'.png',0)\n",
    "                image = windowing(image,200,100)\n",
    "                cv2.imwrite(self.dataPath+'test/'+name+'.png', image)\n",
    "win = windowImg('train')\n",
    "win.run()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false,
    "execution": {
     "iopub.execute_input": "2023-01-04T18:35:32.315877Z",
     "iopub.status.busy": "2023-01-04T18:35:32.314959Z",
     "iopub.status.idle": "2023-01-04T18:35:33.947580Z",
     "shell.execute_reply": "2023-01-04T18:35:33.946515Z",
     "shell.execute_reply.started": "2023-01-04T18:35:32.315842Z"
    },
    "jupyter": {
     "outputs_hidden": false
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "!mkdir /home/aistudio/PaddleSeg/data/mine/Image_Label\n",
    "class Pre(object):\n",
    "    def __init__(self):\n",
    "        self.dataPath = '/home/aistudio/PaddleSeg/data/mine/'\n",
    "    def run(self):\n",
    "        # 生成train.txt和valid.txt\n",
    "        trainFile = open(self.dataPath + 'train.txt', 'w')\n",
    "        vaildFile = open(self.dataPath + 'valid.txt', 'w')\n",
    "        for i in range(1,101):\n",
    "            name = str(i).zfill(4)\n",
    "            if i <= 90:\n",
    "                trainFile.write('image/'+name+'.png Image_Label/'+name+'.png')\n",
    "                trainFile.write('\\n')\n",
    "            else:\n",
    "                vaildFile.write('image/'+name+'.png Image_Label/'+name+'.png')\n",
    "                vaildFile.write('\\n')\n",
    "            # 将标签图片转为标准标准图片\n",
    "            image = cv2.imread(self.dataPath+'Layer_Masks/'+name+'.png')\n",
    "            # 像素值为0的是RNFL(类别 0)\n",
    "            # 像素值为80的是GCIPL(类别 1)\n",
    "            # 像素值为160的是脉络膜(类别 2)\n",
    "            # 像素值为255的是其他（类别3）。\n",
    "            image[image == 80] = 1\n",
    "            image[image == 160] = 2\n",
    "            image[image == 255] = 3\n",
    "            image = image[:,:,1]\n",
    "            cv2.imwrite(self.dataPath+'Image_Label/'+name+'.png', image)\n",
    "pre = Pre()\n",
    "pre.run()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 三.训练\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 模型\n",
    "\n",
    "模型使用了早期的网络U-net。刚刚发现了官方baseline用的也是U net。这另一方面说明了图像类任务主要是预处理和后处理，比赛而言模型大部分时候不会改。\n",
    "\n",
    "* 医学图像的语义较为简单，每一个部分都是人的很小一部分。不同的器官也会用不同的器械。器官本身结构固定，所以对于一个语义分割任务只有那么几类，不像是自动驾驶，室外场景会碰到各种不同的类别。进而高级语义信息和底层特征都很重要。这样就是U net的特点，高层特征和底层特征融合，这样网络兼顾两者。\n",
    "* 数据量少。本数据只有100张图片，模型参数不宜多。\n",
    "\n",
    "U net的结构如下图，是一个U型的结果。左右是encoder，包含若干的卷积、降采样层。右方是decoder上采样，同时会把左面的特征进行拼接。右侧反卷集，最后恢复到图像原始形状。\n",
    "\n",
    "![](https://ai-studio-static-online.cdn.bcebos.com/e38367730dc04f989373528e3be00e8e6315850ee175457497791f69c3dfb1ea)\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "配置文件我已放到家目录下面，后面直接复制到对应路径，覆盖掉原始的配置文件。\n",
    "\n",
    "首先是改了cityscapes,用于读取数据。\n",
    "```\n",
    "batch_size: 4\n",
    "iters: 2000\n",
    "\n",
    "train_dataset:\n",
    "  type: Dataset\n",
    "  dataset_root: data/mine\n",
    "  train_path: data/mine/train.txt\n",
    "  transforms:\n",
    "    - type: ResizeStepScaling\n",
    "      min_scale_factor: 0.5\n",
    "      max_scale_factor: 2.0\n",
    "      scale_step_size: 0.25\n",
    "    - type: RandomPaddingCrop\n",
    "      crop_size: [1024, 512]\n",
    "    - type: RandomHorizontalFlip\n",
    "    - type: RandomDistort\n",
    "      brightness_range: 0.4\n",
    "      contrast_range: 0.4\n",
    "      saturation_range: 0.4\n",
    "    - type: Normalize\n",
    "  num_classes: 4\n",
    "  mode: train\n",
    "\n",
    "val_dataset:\n",
    "  type: Dataset\n",
    "  dataset_root: data/mine\n",
    "  val_path: data/mine/valid.txt\n",
    "  transforms:\n",
    "    - type: Normalize\n",
    "  num_classes: 4\n",
    "  mode: val\n",
    "\n",
    "\n",
    "optimizer:\n",
    "  type: sgd\n",
    "  momentum: 0.9\n",
    "  weight_decay: 4.0e-5\n",
    "\n",
    "lr_scheduler:\n",
    "  type: PolynomialDecay\n",
    "  learning_rate: 0.01\n",
    "  end_lr: 0\n",
    "  power: 0.9\n",
    "\n",
    "loss:\n",
    "  types:\n",
    "    - type: CrossEntropyLoss\n",
    "  coef: [1]\n",
    "```\n",
    "\n",
    "然后是修改了适配的模型yml\n",
    "\n",
    "```\n",
    "_base_: '../_base_/cityscapes.yml'\n",
    "\n",
    "batch_size: 4\n",
    "iters: 2000\n",
    "\n",
    "model:\n",
    "  type: UNet\n",
    "  num_classes: 4\n",
    "  use_deconv: False\n",
    "  pretrained: Null\n",
    "\n",
    "```\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-04T18:32:51.073095Z",
     "iopub.status.busy": "2023-01-04T18:32:51.072418Z",
     "iopub.status.idle": "2023-01-04T18:32:51.530278Z",
     "shell.execute_reply": "2023-01-04T18:32:51.529058Z",
     "shell.execute_reply.started": "2023-01-04T18:32:51.073062Z"
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "!cp /home/aistudio/unet_cityscapes_1024x512_160k.yml /home/aistudio/PaddleSeg/configs/unet\n",
    "!cp /home/aistudio/cityscapes.yml /home/aistudio/PaddleSeg/configs/_base_\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false,
    "execution": {
     "iopub.execute_input": "2023-01-04T18:35:54.483897Z",
     "iopub.status.busy": "2023-01-04T18:35:54.482952Z",
     "iopub.status.idle": "2023-01-04T18:59:02.940071Z",
     "shell.execute_reply": "2023-01-04T18:59:02.939039Z",
     "shell.execute_reply.started": "2023-01-04T18:35:54.483860Z"
    },
    "jupyter": {
     "outputs_hidden": false
    },
    "scrolled": true,
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/home/aistudio/PaddleSeg\r\n",
      "2023-01-05 02:35:56 [INFO]\t\r\n",
      "------------Environment Information-------------\r\n",
      "platform: Linux-4.15.0-140-generic-x86_64-with-debian-stretch-sid\r\n",
      "Python: 3.7.4 (default, Aug 13 2019, 20:35:49) [GCC 7.3.0]\r\n",
      "Paddle compiled with cuda: True\r\n",
      "NVCC: Build cuda_11.2.r11.2/compiler.29618528_0\r\n",
      "cudnn: 8.2\r\n",
      "GPUs used: 1\r\n",
      "CUDA_VISIBLE_DEVICES: None\r\n",
      "GPU: ['GPU 0: Tesla V100-SXM2-16GB']\r\n",
      "GCC: gcc (Ubuntu 7.5.0-3ubuntu1~16.04) 7.5.0\r\n",
      "PaddleSeg: 2.6.0\r\n",
      "PaddlePaddle: 2.3.2\r\n",
      "OpenCV: 4.6.0\r\n",
      "------------------------------------------------\r\n",
      "2023-01-05 02:35:56 [INFO]\t\r\n",
      "---------------Config Information---------------\r\n",
      "batch_size: 4\r\n",
      "iters: 2000\r\n",
      "loss:\r\n",
      "  coef:\r\n",
      "  - 1\r\n",
      "  types:\r\n",
      "  - ignore_index: 255\r\n",
      "    type: CrossEntropyLoss\r\n",
      "lr_scheduler:\r\n",
      "  end_lr: 0\r\n",
      "  learning_rate: 0.01\r\n",
      "  power: 0.9\r\n",
      "  type: PolynomialDecay\r\n",
      "model:\r\n",
      "  num_classes: 4\r\n",
      "  pretrained: null\r\n",
      "  type: UNet\r\n",
      "  use_deconv: false\r\n",
      "optimizer:\r\n",
      "  momentum: 0.9\r\n",
      "  type: sgd\r\n",
      "  weight_decay: 4.0e-05\r\n",
      "train_dataset:\r\n",
      "  dataset_root: data/mine\r\n",
      "  mode: train\r\n",
      "  num_classes: 4\r\n",
      "  train_path: data/mine/train.txt\r\n",
      "  transforms:\r\n",
      "  - max_scale_factor: 2.0\r\n",
      "    min_scale_factor: 0.5\r\n",
      "    scale_step_size: 0.25\r\n",
      "    type: ResizeStepScaling\r\n",
      "  - crop_size:\r\n",
      "    - 1024\r\n",
      "    - 512\r\n",
      "    type: RandomPaddingCrop\r\n",
      "  - type: RandomHorizontalFlip\r\n",
      "  - brightness_range: 0.4\r\n",
      "    contrast_range: 0.4\r\n",
      "    saturation_range: 0.4\r\n",
      "    type: RandomDistort\r\n",
      "  - type: Normalize\r\n",
      "  type: Dataset\r\n",
      "val_dataset:\r\n",
      "  dataset_root: data/mine\r\n",
      "  mode: val\r\n",
      "  num_classes: 4\r\n",
      "  transforms:\r\n",
      "  - type: Normalize\r\n",
      "  type: Dataset\r\n",
      "  val_path: data/mine/valid.txt\r\n",
      "------------------------------------------------\r\n",
      "W0105 02:35:56.760035 17690 gpu_resources.cc:61] Please NOTE: device: 0, GPU Compute Capability: 7.0, Driver API Version: 11.2, Runtime API Version: 11.2\r\n",
      "W0105 02:35:56.760076 17690 gpu_resources.cc:91] device: 0, cuDNN Version: 8.2.\r\n",
      "/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/nn/layer/norm.py:654: UserWarning: When training, we now always track global mean and variance.\r\n",
      "  \"When training, we now always track global mean and variance.\")\r\n",
      "/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/dygraph/math_op_patch.py:278: UserWarning: The dtype of left and right variables are not the same, left dtype is paddle.float32, but right dtype is paddle.int64, the right dtype will convert to paddle.float32\r\n",
      "  format(lhs_dtype, rhs_dtype, lhs_dtype))\r\n",
      "2023-01-05 02:36:05 [INFO]\t[TRAIN] epoch: 1, iter: 10/2000, loss: 0.5051, lr: 0.009959, batch_cost: 0.8043, reader_cost: 0.09496, ips: 4.9734 samples/sec | ETA 00:26:40\r\n",
      "2023-01-05 02:36:12 [INFO]\t[TRAIN] epoch: 1, iter: 20/2000, loss: 0.3699, lr: 0.009914, batch_cost: 0.6787, reader_cost: 0.18881, ips: 5.8939 samples/sec | ETA 00:22:23\r\n",
      "2023-01-05 02:36:19 [INFO]\t[TRAIN] epoch: 2, iter: 30/2000, loss: 0.3017, lr: 0.009869, batch_cost: 0.6910, reader_cost: 0.20085, ips: 5.7887 samples/sec | ETA 00:22:41\r\n",
      "2023-01-05 02:36:26 [INFO]\t[TRAIN] epoch: 2, iter: 40/2000, loss: 0.2557, lr: 0.009824, batch_cost: 0.6789, reader_cost: 0.18913, ips: 5.8922 samples/sec | ETA 00:22:10\r\n",
      "2023-01-05 02:36:33 [INFO]\t[TRAIN] epoch: 3, iter: 50/2000, loss: 0.2683, lr: 0.009779, batch_cost: 0.6996, reader_cost: 0.20882, ips: 5.7174 samples/sec | ETA 00:22:44\r\n",
      "2023-01-05 02:36:40 [INFO]\t[TRAIN] epoch: 3, iter: 60/2000, loss: 0.1753, lr: 0.009734, batch_cost: 0.6694, reader_cost: 0.17781, ips: 5.9754 samples/sec | ETA 00:21:38\r\n",
      "2023-01-05 02:36:46 [INFO]\t[TRAIN] epoch: 4, iter: 70/2000, loss: 0.1701, lr: 0.009689, batch_cost: 0.6827, reader_cost: 0.19177, ips: 5.8592 samples/sec | ETA 00:21:57\r\n",
      "2023-01-05 02:36:53 [INFO]\t[TRAIN] epoch: 4, iter: 80/2000, loss: 0.1382, lr: 0.009644, batch_cost: 0.6830, reader_cost: 0.19147, ips: 5.8569 samples/sec | ETA 00:21:51\r\n",
      "2023-01-05 02:37:00 [INFO]\t[TRAIN] epoch: 5, iter: 90/2000, loss: 0.1305, lr: 0.009599, batch_cost: 0.6993, reader_cost: 0.20574, ips: 5.7197 samples/sec | ETA 00:22:15\r\n",
      "2023-01-05 02:37:07 [INFO]\t[TRAIN] epoch: 5, iter: 100/2000, loss: 0.1538, lr: 0.009553, batch_cost: 0.6955, reader_cost: 0.19943, ips: 5.7509 samples/sec | ETA 00:22:01\r\n",
      "2023-01-05 02:37:14 [INFO]\t[TRAIN] epoch: 5, iter: 110/2000, loss: 0.1723, lr: 0.009508, batch_cost: 0.6850, reader_cost: 0.19045, ips: 5.8393 samples/sec | ETA 00:21:34\r\n",
      "2023-01-05 02:37:21 [INFO]\t[TRAIN] epoch: 6, iter: 120/2000, loss: 0.1597, lr: 0.009463, batch_cost: 0.6884, reader_cost: 0.19581, ips: 5.8105 samples/sec | ETA 00:21:34\r\n",
      "2023-01-05 02:37:28 [INFO]\t[TRAIN] epoch: 6, iter: 130/2000, loss: 0.1270, lr: 0.009418, batch_cost: 0.6771, reader_cost: 0.18409, ips: 5.9080 samples/sec | ETA 00:21:06\r\n",
      "2023-01-05 02:37:35 [INFO]\t[TRAIN] epoch: 7, iter: 140/2000, loss: 0.1106, lr: 0.009372, batch_cost: 0.6958, reader_cost: 0.20098, ips: 5.7485 samples/sec | ETA 00:21:34\r\n",
      "2023-01-05 02:37:42 [INFO]\t[TRAIN] epoch: 7, iter: 150/2000, loss: 0.0918, lr: 0.009327, batch_cost: 0.6799, reader_cost: 0.18669, ips: 5.8834 samples/sec | ETA 00:20:57\r\n",
      "2023-01-05 02:37:48 [INFO]\t[TRAIN] epoch: 8, iter: 160/2000, loss: 0.1700, lr: 0.009282, batch_cost: 0.6927, reader_cost: 0.19865, ips: 5.7742 samples/sec | ETA 00:21:14\r\n",
      "2023-01-05 02:37:55 [INFO]\t[TRAIN] epoch: 8, iter: 170/2000, loss: 0.1017, lr: 0.009236, batch_cost: 0.6944, reader_cost: 0.19881, ips: 5.7606 samples/sec | ETA 00:21:10\r\n",
      "2023-01-05 02:38:02 [INFO]\t[TRAIN] epoch: 9, iter: 180/2000, loss: 0.1125, lr: 0.009191, batch_cost: 0.6908, reader_cost: 0.19609, ips: 5.7905 samples/sec | ETA 00:20:57\r\n",
      "2023-01-05 02:38:09 [INFO]\t[TRAIN] epoch: 9, iter: 190/2000, loss: 0.0957, lr: 0.009145, batch_cost: 0.6765, reader_cost: 0.18200, ips: 5.9126 samples/sec | ETA 00:20:24\r\n",
      "2023-01-05 02:38:16 [INFO]\t[TRAIN] epoch: 10, iter: 200/2000, loss: 0.0891, lr: 0.009100, batch_cost: 0.6768, reader_cost: 0.18088, ips: 5.9102 samples/sec | ETA 00:20:18\r\n",
      "2023-01-05 02:38:23 [INFO]\t[TRAIN] epoch: 10, iter: 210/2000, loss: 0.0840, lr: 0.009054, batch_cost: 0.6943, reader_cost: 0.19820, ips: 5.7611 samples/sec | ETA 00:20:42\r\n",
      "2023-01-05 02:38:30 [INFO]\t[TRAIN] epoch: 10, iter: 220/2000, loss: 0.0726, lr: 0.009009, batch_cost: 0.6775, reader_cost: 0.18483, ips: 5.9043 samples/sec | ETA 00:20:05\r\n",
      "2023-01-05 02:38:36 [INFO]\t[TRAIN] epoch: 11, iter: 230/2000, loss: 0.0907, lr: 0.008963, batch_cost: 0.6932, reader_cost: 0.19552, ips: 5.7706 samples/sec | ETA 00:20:26\r\n",
      "2023-01-05 02:38:43 [INFO]\t[TRAIN] epoch: 11, iter: 240/2000, loss: 0.1020, lr: 0.008918, batch_cost: 0.6801, reader_cost: 0.18658, ips: 5.8816 samples/sec | ETA 00:19:56\r\n",
      "2023-01-05 02:38:50 [INFO]\t[TRAIN] epoch: 12, iter: 250/2000, loss: 0.1064, lr: 0.008872, batch_cost: 0.7082, reader_cost: 0.21076, ips: 5.6485 samples/sec | ETA 00:20:39\r\n",
      "2023-01-05 02:38:57 [INFO]\t[TRAIN] epoch: 12, iter: 260/2000, loss: 0.0832, lr: 0.008827, batch_cost: 0.6934, reader_cost: 0.19663, ips: 5.7683 samples/sec | ETA 00:20:06\r\n",
      "2023-01-05 02:39:04 [INFO]\t[TRAIN] epoch: 13, iter: 270/2000, loss: 0.0702, lr: 0.008781, batch_cost: 0.6938, reader_cost: 0.19982, ips: 5.7654 samples/sec | ETA 00:20:00\r\n",
      "2023-01-05 02:39:11 [INFO]\t[TRAIN] epoch: 13, iter: 280/2000, loss: 0.0702, lr: 0.008735, batch_cost: 0.6826, reader_cost: 0.18751, ips: 5.8596 samples/sec | ETA 00:19:34\r\n",
      "2023-01-05 02:39:18 [INFO]\t[TRAIN] epoch: 14, iter: 290/2000, loss: 0.0751, lr: 0.008690, batch_cost: 0.6959, reader_cost: 0.19997, ips: 5.7477 samples/sec | ETA 00:19:50\r\n",
      "2023-01-05 02:39:25 [INFO]\t[TRAIN] epoch: 14, iter: 300/2000, loss: 0.0870, lr: 0.008644, batch_cost: 0.6908, reader_cost: 0.19283, ips: 5.7904 samples/sec | ETA 00:19:34\r\n",
      "2023-01-05 02:39:32 [INFO]\t[TRAIN] epoch: 15, iter: 310/2000, loss: 0.0618, lr: 0.008598, batch_cost: 0.6919, reader_cost: 0.19595, ips: 5.7811 samples/sec | ETA 00:19:29\r\n",
      "2023-01-05 02:39:39 [INFO]\t[TRAIN] epoch: 15, iter: 320/2000, loss: 0.0660, lr: 0.008552, batch_cost: 0.6876, reader_cost: 0.19288, ips: 5.8172 samples/sec | ETA 00:19:15\r\n",
      "2023-01-05 02:39:46 [INFO]\t[TRAIN] epoch: 15, iter: 330/2000, loss: 0.0621, lr: 0.008507, batch_cost: 0.6965, reader_cost: 0.20068, ips: 5.7427 samples/sec | ETA 00:19:23\r\n",
      "2023-01-05 02:39:53 [INFO]\t[TRAIN] epoch: 16, iter: 340/2000, loss: 0.0673, lr: 0.008461, batch_cost: 0.6902, reader_cost: 0.19240, ips: 5.7952 samples/sec | ETA 00:19:05\r\n",
      "2023-01-05 02:39:59 [INFO]\t[TRAIN] epoch: 16, iter: 350/2000, loss: 0.1199, lr: 0.008415, batch_cost: 0.6839, reader_cost: 0.18928, ips: 5.8491 samples/sec | ETA 00:18:48\r\n",
      "2023-01-05 02:40:06 [INFO]\t[TRAIN] epoch: 17, iter: 360/2000, loss: 0.0616, lr: 0.008369, batch_cost: 0.6945, reader_cost: 0.20026, ips: 5.7599 samples/sec | ETA 00:18:58\r\n",
      "2023-01-05 02:40:13 [INFO]\t[TRAIN] epoch: 17, iter: 370/2000, loss: 0.1074, lr: 0.008323, batch_cost: 0.6935, reader_cost: 0.19678, ips: 5.7680 samples/sec | ETA 00:18:50\r\n",
      "2023-01-05 02:40:20 [INFO]\t[TRAIN] epoch: 18, iter: 380/2000, loss: 0.1001, lr: 0.008277, batch_cost: 0.7001, reader_cost: 0.20230, ips: 5.7133 samples/sec | ETA 00:18:54\r\n",
      "2023-01-05 02:40:27 [INFO]\t[TRAIN] epoch: 18, iter: 390/2000, loss: 0.0883, lr: 0.008231, batch_cost: 0.6741, reader_cost: 0.17990, ips: 5.9340 samples/sec | ETA 00:18:05\r\n",
      "2023-01-05 02:40:34 [INFO]\t[TRAIN] epoch: 19, iter: 400/2000, loss: 0.0674, lr: 0.008185, batch_cost: 0.6900, reader_cost: 0.19355, ips: 5.7969 samples/sec | ETA 00:18:24\r\n",
      "2023-01-05 02:40:41 [INFO]\t[TRAIN] epoch: 19, iter: 410/2000, loss: 0.0721, lr: 0.008139, batch_cost: 0.6754, reader_cost: 0.18022, ips: 5.9228 samples/sec | ETA 00:17:53\r\n",
      "2023-01-05 02:40:48 [INFO]\t[TRAIN] epoch: 20, iter: 420/2000, loss: 0.0801, lr: 0.008093, batch_cost: 0.6945, reader_cost: 0.19685, ips: 5.7593 samples/sec | ETA 00:18:17\r\n",
      "2023-01-05 02:40:55 [INFO]\t[TRAIN] epoch: 20, iter: 430/2000, loss: 0.0802, lr: 0.008047, batch_cost: 0.6918, reader_cost: 0.19511, ips: 5.7819 samples/sec | ETA 00:18:06\r\n",
      "2023-01-05 02:41:01 [INFO]\t[TRAIN] epoch: 20, iter: 440/2000, loss: 0.0571, lr: 0.008001, batch_cost: 0.6895, reader_cost: 0.19277, ips: 5.8012 samples/sec | ETA 00:17:55\r\n",
      "2023-01-05 02:41:08 [INFO]\t[TRAIN] epoch: 21, iter: 450/2000, loss: 0.0864, lr: 0.007955, batch_cost: 0.6992, reader_cost: 0.20084, ips: 5.7204 samples/sec | ETA 00:18:03\r\n",
      "2023-01-05 02:41:15 [INFO]\t[TRAIN] epoch: 21, iter: 460/2000, loss: 0.0736, lr: 0.007909, batch_cost: 0.6832, reader_cost: 0.18782, ips: 5.8550 samples/sec | ETA 00:17:32\r\n",
      "2023-01-05 02:41:22 [INFO]\t[TRAIN] epoch: 22, iter: 470/2000, loss: 0.0613, lr: 0.007862, batch_cost: 0.7103, reader_cost: 0.21395, ips: 5.6313 samples/sec | ETA 00:18:06\r\n",
      "2023-01-05 02:41:29 [INFO]\t[TRAIN] epoch: 22, iter: 480/2000, loss: 0.0590, lr: 0.007816, batch_cost: 0.6846, reader_cost: 0.18790, ips: 5.8427 samples/sec | ETA 00:17:20\r\n",
      "2023-01-05 02:41:36 [INFO]\t[TRAIN] epoch: 23, iter: 490/2000, loss: 0.0771, lr: 0.007770, batch_cost: 0.6818, reader_cost: 0.18500, ips: 5.8668 samples/sec | ETA 00:17:09\r\n",
      "2023-01-05 02:41:43 [INFO]\t[TRAIN] epoch: 23, iter: 500/2000, loss: 0.0691, lr: 0.007724, batch_cost: 0.6955, reader_cost: 0.19759, ips: 5.7511 samples/sec | ETA 00:17:23\r\n",
      "2023-01-05 02:41:43 [INFO]\tStart evaluating (total_samples: 10, total_iters: 10)...\r\n",
      "10/10 [==============================] - 1s 79ms/step - batch_cost: 0.0783 - reader cost: 0.0423\r\n",
      "2023-01-05 02:41:44 [INFO]\t[EVAL] #Images: 10 mIoU: 0.8580 Acc: 0.9851 Kappa: 0.9098 Dice: 0.9216\r\n",
      "2023-01-05 02:41:44 [INFO]\t[EVAL] Class IoU: \r\n",
      "[0.8545 0.7737 0.8171 0.9866]\r\n",
      "2023-01-05 02:41:44 [INFO]\t[EVAL] Class Precision: \r\n",
      "[0.963  0.8582 0.9338 0.9906]\r\n",
      "2023-01-05 02:41:44 [INFO]\t[EVAL] Class Recall: \r\n",
      "[0.8835 0.8872 0.8673 0.9959]\r\n",
      "2023-01-05 02:41:44 [INFO]\t[EVAL] The model with the best validation mIoU (0.8580) was saved at iter 500.\r\n",
      "2023-01-05 02:41:50 [INFO]\t[TRAIN] epoch: 24, iter: 510/2000, loss: 0.0648, lr: 0.007677, batch_cost: 0.6092, reader_cost: 0.11218, ips: 6.5658 samples/sec | ETA 00:15:07\r\n",
      "2023-01-05 02:41:57 [INFO]\t[TRAIN] epoch: 24, iter: 520/2000, loss: 0.0728, lr: 0.007631, batch_cost: 0.6875, reader_cost: 0.18962, ips: 5.8178 samples/sec | ETA 00:16:57\r\n",
      "2023-01-05 02:42:04 [INFO]\t[TRAIN] epoch: 25, iter: 530/2000, loss: 0.0689, lr: 0.007584, batch_cost: 0.6950, reader_cost: 0.19710, ips: 5.7557 samples/sec | ETA 00:17:01\r\n",
      "2023-01-05 02:42:11 [INFO]\t[TRAIN] epoch: 25, iter: 540/2000, loss: 0.0852, lr: 0.007538, batch_cost: 0.6910, reader_cost: 0.19169, ips: 5.7889 samples/sec | ETA 00:16:48\r\n",
      "2023-01-05 02:42:18 [INFO]\t[TRAIN] epoch: 25, iter: 550/2000, loss: 0.1154, lr: 0.007492, batch_cost: 0.6778, reader_cost: 0.17957, ips: 5.9011 samples/sec | ETA 00:16:22\r\n",
      "2023-01-05 02:42:25 [INFO]\t[TRAIN] epoch: 26, iter: 560/2000, loss: 0.0815, lr: 0.007445, batch_cost: 0.6999, reader_cost: 0.20359, ips: 5.7147 samples/sec | ETA 00:16:47\r\n",
      "2023-01-05 02:42:32 [INFO]\t[TRAIN] epoch: 26, iter: 570/2000, loss: 0.0677, lr: 0.007399, batch_cost: 0.6939, reader_cost: 0.19672, ips: 5.7645 samples/sec | ETA 00:16:32\r\n",
      "2023-01-05 02:42:39 [INFO]\t[TRAIN] epoch: 27, iter: 580/2000, loss: 0.0687, lr: 0.007352, batch_cost: 0.6870, reader_cost: 0.19029, ips: 5.8227 samples/sec | ETA 00:16:15\r\n",
      "2023-01-05 02:42:45 [INFO]\t[TRAIN] epoch: 27, iter: 590/2000, loss: 0.0549, lr: 0.007305, batch_cost: 0.6720, reader_cost: 0.17397, ips: 5.9526 samples/sec | ETA 00:15:47\r\n",
      "2023-01-05 02:42:52 [INFO]\t[TRAIN] epoch: 28, iter: 600/2000, loss: 0.0785, lr: 0.007259, batch_cost: 0.7040, reader_cost: 0.20751, ips: 5.6816 samples/sec | ETA 00:16:25\r\n",
      "2023-01-05 02:42:59 [INFO]\t[TRAIN] epoch: 28, iter: 610/2000, loss: 0.0563, lr: 0.007212, batch_cost: 0.7033, reader_cost: 0.20602, ips: 5.6874 samples/sec | ETA 00:16:17\r\n",
      "2023-01-05 02:43:06 [INFO]\t[TRAIN] epoch: 29, iter: 620/2000, loss: 0.0521, lr: 0.007166, batch_cost: 0.6965, reader_cost: 0.20033, ips: 5.7433 samples/sec | ETA 00:16:01\r\n",
      "2023-01-05 02:43:13 [INFO]\t[TRAIN] epoch: 29, iter: 630/2000, loss: 0.0671, lr: 0.007119, batch_cost: 0.6777, reader_cost: 0.18012, ips: 5.9021 samples/sec | ETA 00:15:28\r\n",
      "2023-01-05 02:43:20 [INFO]\t[TRAIN] epoch: 30, iter: 640/2000, loss: 0.0584, lr: 0.007072, batch_cost: 0.6905, reader_cost: 0.19080, ips: 5.7926 samples/sec | ETA 00:15:39\r\n",
      "2023-01-05 02:43:27 [INFO]\t[TRAIN] epoch: 30, iter: 650/2000, loss: 0.0776, lr: 0.007025, batch_cost: 0.6924, reader_cost: 0.19612, ips: 5.7768 samples/sec | ETA 00:15:34\r\n",
      "2023-01-05 02:43:34 [INFO]\t[TRAIN] epoch: 30, iter: 660/2000, loss: 0.0518, lr: 0.006978, batch_cost: 0.6915, reader_cost: 0.19463, ips: 5.7844 samples/sec | ETA 00:15:26\r\n",
      "2023-01-05 02:43:41 [INFO]\t[TRAIN] epoch: 31, iter: 670/2000, loss: 0.0556, lr: 0.006932, batch_cost: 0.6872, reader_cost: 0.19094, ips: 5.8206 samples/sec | ETA 00:15:13\r\n",
      "2023-01-05 02:43:48 [INFO]\t[TRAIN] epoch: 31, iter: 680/2000, loss: 0.0504, lr: 0.006885, batch_cost: 0.6864, reader_cost: 0.18818, ips: 5.8275 samples/sec | ETA 00:15:06\r\n",
      "2023-01-05 02:43:55 [INFO]\t[TRAIN] epoch: 32, iter: 690/2000, loss: 0.0548, lr: 0.006838, batch_cost: 0.6900, reader_cost: 0.19259, ips: 5.7968 samples/sec | ETA 00:15:03\r\n",
      "2023-01-05 02:44:01 [INFO]\t[TRAIN] epoch: 32, iter: 700/2000, loss: 0.0540, lr: 0.006791, batch_cost: 0.6877, reader_cost: 0.18675, ips: 5.8164 samples/sec | ETA 00:14:54\r\n",
      "2023-01-05 02:44:08 [INFO]\t[TRAIN] epoch: 33, iter: 710/2000, loss: 0.0651, lr: 0.006744, batch_cost: 0.6855, reader_cost: 0.18755, ips: 5.8353 samples/sec | ETA 00:14:44\r\n",
      "2023-01-05 02:44:15 [INFO]\t[TRAIN] epoch: 33, iter: 720/2000, loss: 0.0473, lr: 0.006697, batch_cost: 0.6842, reader_cost: 0.18650, ips: 5.8461 samples/sec | ETA 00:14:35\r\n",
      "2023-01-05 02:44:22 [INFO]\t[TRAIN] epoch: 34, iter: 730/2000, loss: 0.0631, lr: 0.006650, batch_cost: 0.6941, reader_cost: 0.19468, ips: 5.7628 samples/sec | ETA 00:14:41\r\n",
      "2023-01-05 02:44:29 [INFO]\t[TRAIN] epoch: 34, iter: 740/2000, loss: 0.0595, lr: 0.006603, batch_cost: 0.6915, reader_cost: 0.19471, ips: 5.7844 samples/sec | ETA 00:14:31\r\n",
      "2023-01-05 02:44:36 [INFO]\t[TRAIN] epoch: 35, iter: 750/2000, loss: 0.0586, lr: 0.006555, batch_cost: 0.7011, reader_cost: 0.20382, ips: 5.7051 samples/sec | ETA 00:14:36\r\n",
      "2023-01-05 02:44:43 [INFO]\t[TRAIN] epoch: 35, iter: 760/2000, loss: 0.0509, lr: 0.006508, batch_cost: 0.6993, reader_cost: 0.20055, ips: 5.7198 samples/sec | ETA 00:14:27\r\n",
      "2023-01-05 02:44:50 [INFO]\t[TRAIN] epoch: 35, iter: 770/2000, loss: 0.0708, lr: 0.006461, batch_cost: 0.6853, reader_cost: 0.18950, ips: 5.8370 samples/sec | ETA 00:14:02\r\n",
      "2023-01-05 02:44:57 [INFO]\t[TRAIN] epoch: 36, iter: 780/2000, loss: 0.0586, lr: 0.006414, batch_cost: 0.7145, reader_cost: 0.21793, ips: 5.5984 samples/sec | ETA 00:14:31\r\n",
      "2023-01-05 02:45:04 [INFO]\t[TRAIN] epoch: 36, iter: 790/2000, loss: 0.0470, lr: 0.006367, batch_cost: 0.6854, reader_cost: 0.18682, ips: 5.8360 samples/sec | ETA 00:13:49\r\n",
      "2023-01-05 02:45:11 [INFO]\t[TRAIN] epoch: 37, iter: 800/2000, loss: 0.0786, lr: 0.006319, batch_cost: 0.6893, reader_cost: 0.19121, ips: 5.8031 samples/sec | ETA 00:13:47\r\n",
      "2023-01-05 02:45:18 [INFO]\t[TRAIN] epoch: 37, iter: 810/2000, loss: 0.0437, lr: 0.006272, batch_cost: 0.6761, reader_cost: 0.18067, ips: 5.9166 samples/sec | ETA 00:13:24\r\n",
      "2023-01-05 02:45:24 [INFO]\t[TRAIN] epoch: 38, iter: 820/2000, loss: 0.0595, lr: 0.006224, batch_cost: 0.6981, reader_cost: 0.20074, ips: 5.7298 samples/sec | ETA 00:13:43\r\n",
      "2023-01-05 02:45:31 [INFO]\t[TRAIN] epoch: 38, iter: 830/2000, loss: 0.0582, lr: 0.006177, batch_cost: 0.6904, reader_cost: 0.19336, ips: 5.7934 samples/sec | ETA 00:13:27\r\n",
      "2023-01-05 02:45:38 [INFO]\t[TRAIN] epoch: 39, iter: 840/2000, loss: 0.0773, lr: 0.006129, batch_cost: 0.6872, reader_cost: 0.18986, ips: 5.8206 samples/sec | ETA 00:13:17\r\n",
      "2023-01-05 02:45:45 [INFO]\t[TRAIN] epoch: 39, iter: 850/2000, loss: 0.0713, lr: 0.006082, batch_cost: 0.6710, reader_cost: 0.17616, ips: 5.9608 samples/sec | ETA 00:12:51\r\n",
      "2023-01-05 02:45:52 [INFO]\t[TRAIN] epoch: 40, iter: 860/2000, loss: 0.0553, lr: 0.006034, batch_cost: 0.7041, reader_cost: 0.20701, ips: 5.6812 samples/sec | ETA 00:13:22\r\n",
      "2023-01-05 02:45:59 [INFO]\t[TRAIN] epoch: 40, iter: 870/2000, loss: 0.0506, lr: 0.005987, batch_cost: 0.6861, reader_cost: 0.18665, ips: 5.8305 samples/sec | ETA 00:12:55\r\n",
      "2023-01-05 02:46:06 [INFO]\t[TRAIN] epoch: 40, iter: 880/2000, loss: 0.0522, lr: 0.005939, batch_cost: 0.6963, reader_cost: 0.19635, ips: 5.7444 samples/sec | ETA 00:12:59\r\n",
      "2023-01-05 02:46:13 [INFO]\t[TRAIN] epoch: 41, iter: 890/2000, loss: 0.0438, lr: 0.005891, batch_cost: 0.6809, reader_cost: 0.18368, ips: 5.8744 samples/sec | ETA 00:12:35\r\n",
      "2023-01-05 02:46:19 [INFO]\t[TRAIN] epoch: 41, iter: 900/2000, loss: 0.0743, lr: 0.005844, batch_cost: 0.6771, reader_cost: 0.18041, ips: 5.9075 samples/sec | ETA 00:12:24\r\n",
      "2023-01-05 02:46:26 [INFO]\t[TRAIN] epoch: 42, iter: 910/2000, loss: 0.0442, lr: 0.005796, batch_cost: 0.6856, reader_cost: 0.19066, ips: 5.8343 samples/sec | ETA 00:12:27\r\n",
      "2023-01-05 02:46:33 [INFO]\t[TRAIN] epoch: 42, iter: 920/2000, loss: 0.0645, lr: 0.005748, batch_cost: 0.6900, reader_cost: 0.19217, ips: 5.7967 samples/sec | ETA 00:12:25\r\n",
      "2023-01-05 02:46:40 [INFO]\t[TRAIN] epoch: 43, iter: 930/2000, loss: 0.0612, lr: 0.005700, batch_cost: 0.6873, reader_cost: 0.19050, ips: 5.8202 samples/sec | ETA 00:12:15\r\n",
      "2023-01-05 02:46:47 [INFO]\t[TRAIN] epoch: 43, iter: 940/2000, loss: 0.0594, lr: 0.005652, batch_cost: 0.6898, reader_cost: 0.19247, ips: 5.7987 samples/sec | ETA 00:12:11\r\n",
      "2023-01-05 02:46:54 [INFO]\t[TRAIN] epoch: 44, iter: 950/2000, loss: 0.0687, lr: 0.005604, batch_cost: 0.6967, reader_cost: 0.19857, ips: 5.7410 samples/sec | ETA 00:12:11\r\n",
      "2023-01-05 02:47:01 [INFO]\t[TRAIN] epoch: 44, iter: 960/2000, loss: 0.0508, lr: 0.005556, batch_cost: 0.6778, reader_cost: 0.18279, ips: 5.9011 samples/sec | ETA 00:11:44\r\n",
      "2023-01-05 02:47:08 [INFO]\t[TRAIN] epoch: 45, iter: 970/2000, loss: 0.0474, lr: 0.005508, batch_cost: 0.7176, reader_cost: 0.22062, ips: 5.5740 samples/sec | ETA 00:12:19\r\n",
      "2023-01-05 02:47:15 [INFO]\t[TRAIN] epoch: 45, iter: 980/2000, loss: 0.0441, lr: 0.005460, batch_cost: 0.6895, reader_cost: 0.18948, ips: 5.8014 samples/sec | ETA 00:11:43\r\n",
      "2023-01-05 02:47:22 [INFO]\t[TRAIN] epoch: 45, iter: 990/2000, loss: 0.0728, lr: 0.005412, batch_cost: 0.6806, reader_cost: 0.18380, ips: 5.8775 samples/sec | ETA 00:11:27\r\n",
      "2023-01-05 02:47:28 [INFO]\t[TRAIN] epoch: 46, iter: 1000/2000, loss: 0.0656, lr: 0.005364, batch_cost: 0.6913, reader_cost: 0.19422, ips: 5.7864 samples/sec | ETA 00:11:31\r\n",
      "2023-01-05 02:47:28 [INFO]\tStart evaluating (total_samples: 10, total_iters: 10)...\r\n",
      "10/10 [==============================] - 1s 80ms/step - batch_cost: 0.0798 - reader cost: 0.0378\r\n",
      "2023-01-05 02:47:29 [INFO]\t[EVAL] #Images: 10 mIoU: 0.8771 Acc: 0.9868 Kappa: 0.9239 Dice: 0.9329\r\n",
      "2023-01-05 02:47:29 [INFO]\t[EVAL] Class IoU: \r\n",
      "[0.8838 0.7869 0.8496 0.988 ]\r\n",
      "2023-01-05 02:47:29 [INFO]\t[EVAL] Class Precision: \r\n",
      "[0.9407 0.8506 0.8879 0.9964]\r\n",
      "2023-01-05 02:47:29 [INFO]\t[EVAL] Class Recall: \r\n",
      "[0.936  0.9132 0.9516 0.9916]\r\n",
      "2023-01-05 02:47:30 [INFO]\t[EVAL] The model with the best validation mIoU (0.8771) was saved at iter 1000.\r\n",
      "2023-01-05 02:47:36 [INFO]\t[TRAIN] epoch: 46, iter: 1010/2000, loss: 0.0783, lr: 0.005315, batch_cost: 0.5903, reader_cost: 0.09118, ips: 6.7763 samples/sec | ETA 00:09:44\r\n",
      "2023-01-05 02:47:43 [INFO]\t[TRAIN] epoch: 47, iter: 1020/2000, loss: 0.0802, lr: 0.005267, batch_cost: 0.6885, reader_cost: 0.19053, ips: 5.8094 samples/sec | ETA 00:11:14\r\n",
      "2023-01-05 02:47:50 [INFO]\t[TRAIN] epoch: 47, iter: 1030/2000, loss: 0.0718, lr: 0.005219, batch_cost: 0.6837, reader_cost: 0.18698, ips: 5.8506 samples/sec | ETA 00:11:03\r\n",
      "2023-01-05 02:47:56 [INFO]\t[TRAIN] epoch: 48, iter: 1040/2000, loss: 0.0517, lr: 0.005170, batch_cost: 0.6908, reader_cost: 0.19362, ips: 5.7903 samples/sec | ETA 00:11:03\r\n",
      "2023-01-05 02:48:03 [INFO]\t[TRAIN] epoch: 48, iter: 1050/2000, loss: 0.0555, lr: 0.005122, batch_cost: 0.6852, reader_cost: 0.18802, ips: 5.8377 samples/sec | ETA 00:10:50\r\n",
      "2023-01-05 02:48:10 [INFO]\t[TRAIN] epoch: 49, iter: 1060/2000, loss: 0.0536, lr: 0.005073, batch_cost: 0.7127, reader_cost: 0.21544, ips: 5.6128 samples/sec | ETA 00:11:09\r\n",
      "2023-01-05 02:48:17 [INFO]\t[TRAIN] epoch: 49, iter: 1070/2000, loss: 0.0455, lr: 0.005025, batch_cost: 0.6711, reader_cost: 0.17439, ips: 5.9603 samples/sec | ETA 00:10:24\r\n",
      "2023-01-05 02:48:24 [INFO]\t[TRAIN] epoch: 50, iter: 1080/2000, loss: 0.0599, lr: 0.004976, batch_cost: 0.6924, reader_cost: 0.19662, ips: 5.7771 samples/sec | ETA 00:10:36\r\n",
      "2023-01-05 02:48:31 [INFO]\t[TRAIN] epoch: 50, iter: 1090/2000, loss: 0.0515, lr: 0.004928, batch_cost: 0.7004, reader_cost: 0.20324, ips: 5.7113 samples/sec | ETA 00:10:37\r\n",
      "2023-01-05 02:48:38 [INFO]\t[TRAIN] epoch: 50, iter: 1100/2000, loss: 0.0432, lr: 0.004879, batch_cost: 0.6839, reader_cost: 0.18438, ips: 5.8487 samples/sec | ETA 00:10:15\r\n",
      "2023-01-05 02:48:45 [INFO]\t[TRAIN] epoch: 51, iter: 1110/2000, loss: 0.0410, lr: 0.004830, batch_cost: 0.7018, reader_cost: 0.20403, ips: 5.6998 samples/sec | ETA 00:10:24\r\n",
      "2023-01-05 02:48:52 [INFO]\t[TRAIN] epoch: 51, iter: 1120/2000, loss: 0.0495, lr: 0.004781, batch_cost: 0.6905, reader_cost: 0.19240, ips: 5.7932 samples/sec | ETA 00:10:07\r\n",
      "2023-01-05 02:48:59 [INFO]\t[TRAIN] epoch: 52, iter: 1130/2000, loss: 0.0493, lr: 0.004732, batch_cost: 0.6854, reader_cost: 0.18883, ips: 5.8363 samples/sec | ETA 00:09:56\r\n",
      "2023-01-05 02:49:06 [INFO]\t[TRAIN] epoch: 52, iter: 1140/2000, loss: 0.0418, lr: 0.004684, batch_cost: 0.6831, reader_cost: 0.18509, ips: 5.8558 samples/sec | ETA 00:09:47\r\n",
      "2023-01-05 02:49:12 [INFO]\t[TRAIN] epoch: 53, iter: 1150/2000, loss: 0.0561, lr: 0.004635, batch_cost: 0.6940, reader_cost: 0.19657, ips: 5.7636 samples/sec | ETA 00:09:49\r\n",
      "2023-01-05 02:49:19 [INFO]\t[TRAIN] epoch: 53, iter: 1160/2000, loss: 0.0536, lr: 0.004586, batch_cost: 0.6745, reader_cost: 0.17636, ips: 5.9303 samples/sec | ETA 00:09:26\r\n",
      "2023-01-05 02:49:26 [INFO]\t[TRAIN] epoch: 54, iter: 1170/2000, loss: 0.0456, lr: 0.004536, batch_cost: 0.6922, reader_cost: 0.19432, ips: 5.7784 samples/sec | ETA 00:09:34\r\n",
      "2023-01-05 02:49:33 [INFO]\t[TRAIN] epoch: 54, iter: 1180/2000, loss: 0.0542, lr: 0.004487, batch_cost: 0.6912, reader_cost: 0.19253, ips: 5.7870 samples/sec | ETA 00:09:26\r\n",
      "2023-01-05 02:49:40 [INFO]\t[TRAIN] epoch: 55, iter: 1190/2000, loss: 0.0476, lr: 0.004438, batch_cost: 0.6847, reader_cost: 0.18773, ips: 5.8417 samples/sec | ETA 00:09:14\r\n",
      "2023-01-05 02:49:47 [INFO]\t[TRAIN] epoch: 55, iter: 1200/2000, loss: 0.0516, lr: 0.004389, batch_cost: 0.6938, reader_cost: 0.19508, ips: 5.7651 samples/sec | ETA 00:09:15\r\n",
      "2023-01-05 02:49:54 [INFO]\t[TRAIN] epoch: 55, iter: 1210/2000, loss: 0.0544, lr: 0.004339, batch_cost: 0.6848, reader_cost: 0.18588, ips: 5.8415 samples/sec | ETA 00:09:00\r\n",
      "2023-01-05 02:50:01 [INFO]\t[TRAIN] epoch: 56, iter: 1220/2000, loss: 0.0446, lr: 0.004290, batch_cost: 0.6942, reader_cost: 0.19631, ips: 5.7623 samples/sec | ETA 00:09:01\r\n",
      "2023-01-05 02:50:08 [INFO]\t[TRAIN] epoch: 56, iter: 1230/2000, loss: 0.0621, lr: 0.004241, batch_cost: 0.6925, reader_cost: 0.19384, ips: 5.7762 samples/sec | ETA 00:08:53\r\n",
      "2023-01-05 02:50:15 [INFO]\t[TRAIN] epoch: 57, iter: 1240/2000, loss: 0.0495, lr: 0.004191, batch_cost: 0.7053, reader_cost: 0.20479, ips: 5.6716 samples/sec | ETA 00:08:56\r\n",
      "2023-01-05 02:50:21 [INFO]\t[TRAIN] epoch: 57, iter: 1250/2000, loss: 0.0433, lr: 0.004141, batch_cost: 0.6828, reader_cost: 0.18554, ips: 5.8584 samples/sec | ETA 00:08:32\r\n",
      "2023-01-05 02:50:28 [INFO]\t[TRAIN] epoch: 58, iter: 1260/2000, loss: 0.0470, lr: 0.004092, batch_cost: 0.6821, reader_cost: 0.18602, ips: 5.8644 samples/sec | ETA 00:08:24\r\n",
      "2023-01-05 02:50:35 [INFO]\t[TRAIN] epoch: 58, iter: 1270/2000, loss: 0.0511, lr: 0.004042, batch_cost: 0.6773, reader_cost: 0.18025, ips: 5.9062 samples/sec | ETA 00:08:14\r\n",
      "2023-01-05 02:50:42 [INFO]\t[TRAIN] epoch: 59, iter: 1280/2000, loss: 0.0408, lr: 0.003992, batch_cost: 0.6833, reader_cost: 0.18594, ips: 5.8540 samples/sec | ETA 00:08:11\r\n",
      "2023-01-05 02:50:49 [INFO]\t[TRAIN] epoch: 59, iter: 1290/2000, loss: 0.0487, lr: 0.003942, batch_cost: 0.6987, reader_cost: 0.19948, ips: 5.7252 samples/sec | ETA 00:08:16\r\n",
      "2023-01-05 02:50:56 [INFO]\t[TRAIN] epoch: 60, iter: 1300/2000, loss: 0.0457, lr: 0.003892, batch_cost: 0.6880, reader_cost: 0.19031, ips: 5.8137 samples/sec | ETA 00:08:01\r\n",
      "2023-01-05 02:51:03 [INFO]\t[TRAIN] epoch: 60, iter: 1310/2000, loss: 0.0444, lr: 0.003842, batch_cost: 0.6806, reader_cost: 0.18248, ips: 5.8770 samples/sec | ETA 00:07:49\r\n",
      "2023-01-05 02:51:09 [INFO]\t[TRAIN] epoch: 60, iter: 1320/2000, loss: 0.0377, lr: 0.003792, batch_cost: 0.6806, reader_cost: 0.18339, ips: 5.8774 samples/sec | ETA 00:07:42\r\n",
      "2023-01-05 02:51:16 [INFO]\t[TRAIN] epoch: 61, iter: 1330/2000, loss: 0.0481, lr: 0.003742, batch_cost: 0.6811, reader_cost: 0.18373, ips: 5.8728 samples/sec | ETA 00:07:36\r\n",
      "2023-01-05 02:51:23 [INFO]\t[TRAIN] epoch: 61, iter: 1340/2000, loss: 0.0408, lr: 0.003692, batch_cost: 0.6825, reader_cost: 0.18145, ips: 5.8605 samples/sec | ETA 00:07:30\r\n",
      "2023-01-05 02:51:30 [INFO]\t[TRAIN] epoch: 62, iter: 1350/2000, loss: 0.0414, lr: 0.003642, batch_cost: 0.6904, reader_cost: 0.19434, ips: 5.7941 samples/sec | ETA 00:07:28\r\n",
      "2023-01-05 02:51:37 [INFO]\t[TRAIN] epoch: 62, iter: 1360/2000, loss: 0.0483, lr: 0.003591, batch_cost: 0.6932, reader_cost: 0.19574, ips: 5.7706 samples/sec | ETA 00:07:23\r\n",
      "2023-01-05 02:51:44 [INFO]\t[TRAIN] epoch: 63, iter: 1370/2000, loss: 0.0431, lr: 0.003541, batch_cost: 0.6938, reader_cost: 0.19563, ips: 5.7652 samples/sec | ETA 00:07:17\r\n",
      "2023-01-05 02:51:51 [INFO]\t[TRAIN] epoch: 63, iter: 1380/2000, loss: 0.0396, lr: 0.003490, batch_cost: 0.6916, reader_cost: 0.19390, ips: 5.7839 samples/sec | ETA 00:07:08\r\n",
      "2023-01-05 02:51:58 [INFO]\t[TRAIN] epoch: 64, iter: 1390/2000, loss: 0.0452, lr: 0.003440, batch_cost: 0.6910, reader_cost: 0.19296, ips: 5.7886 samples/sec | ETA 00:07:01\r\n",
      "2023-01-05 02:52:05 [INFO]\t[TRAIN] epoch: 64, iter: 1400/2000, loss: 0.0477, lr: 0.003389, batch_cost: 0.7073, reader_cost: 0.20893, ips: 5.6555 samples/sec | ETA 00:07:04\r\n",
      "2023-01-05 02:52:12 [INFO]\t[TRAIN] epoch: 65, iter: 1410/2000, loss: 0.0450, lr: 0.003338, batch_cost: 0.7182, reader_cost: 0.22038, ips: 5.5694 samples/sec | ETA 00:07:03\r\n",
      "2023-01-05 02:52:19 [INFO]\t[TRAIN] epoch: 65, iter: 1420/2000, loss: 0.0471, lr: 0.003287, batch_cost: 0.6848, reader_cost: 0.18762, ips: 5.8413 samples/sec | ETA 00:06:37\r\n",
      "2023-01-05 02:52:26 [INFO]\t[TRAIN] epoch: 65, iter: 1430/2000, loss: 0.0513, lr: 0.003236, batch_cost: 0.6946, reader_cost: 0.19623, ips: 5.7591 samples/sec | ETA 00:06:35\r\n",
      "2023-01-05 02:52:33 [INFO]\t[TRAIN] epoch: 66, iter: 1440/2000, loss: 0.0405, lr: 0.003185, batch_cost: 0.6941, reader_cost: 0.19798, ips: 5.7631 samples/sec | ETA 00:06:28\r\n",
      "2023-01-05 02:52:39 [INFO]\t[TRAIN] epoch: 66, iter: 1450/2000, loss: 0.0507, lr: 0.003134, batch_cost: 0.6826, reader_cost: 0.18676, ips: 5.8595 samples/sec | ETA 00:06:15\r\n",
      "2023-01-05 02:52:46 [INFO]\t[TRAIN] epoch: 67, iter: 1460/2000, loss: 0.0439, lr: 0.003083, batch_cost: 0.7012, reader_cost: 0.20486, ips: 5.7048 samples/sec | ETA 00:06:18\r\n",
      "2023-01-05 02:52:53 [INFO]\t[TRAIN] epoch: 67, iter: 1470/2000, loss: 0.0441, lr: 0.003032, batch_cost: 0.6925, reader_cost: 0.19511, ips: 5.7759 samples/sec | ETA 00:06:07\r\n",
      "2023-01-05 02:53:00 [INFO]\t[TRAIN] epoch: 68, iter: 1480/2000, loss: 0.0427, lr: 0.002980, batch_cost: 0.6846, reader_cost: 0.18785, ips: 5.8432 samples/sec | ETA 00:05:55\r\n",
      "2023-01-05 02:53:07 [INFO]\t[TRAIN] epoch: 68, iter: 1490/2000, loss: 0.0464, lr: 0.002929, batch_cost: 0.6790, reader_cost: 0.18400, ips: 5.8908 samples/sec | ETA 00:05:46\r\n",
      "2023-01-05 02:53:14 [INFO]\t[TRAIN] epoch: 69, iter: 1500/2000, loss: 0.0384, lr: 0.002877, batch_cost: 0.6973, reader_cost: 0.19764, ips: 5.7364 samples/sec | ETA 00:05:48\r\n",
      "2023-01-05 02:53:14 [INFO]\tStart evaluating (total_samples: 10, total_iters: 10)...\r\n",
      "10/10 [==============================] - 1s 81ms/step - batch_cost: 0.0804 - reader cost: 0.0416\r\n",
      "2023-01-05 02:53:15 [INFO]\t[EVAL] #Images: 10 mIoU: 0.8813 Acc: 0.9880 Kappa: 0.9292 Dice: 0.9354\r\n",
      "2023-01-05 02:53:15 [INFO]\t[EVAL] Class IoU: \r\n",
      "[0.8836 0.7916 0.8608 0.9893]\r\n",
      "2023-01-05 02:53:15 [INFO]\t[EVAL] Class Precision: \r\n",
      "[0.9242 0.9118 0.9165 0.995 ]\r\n",
      "2023-01-05 02:53:15 [INFO]\t[EVAL] Class Recall: \r\n",
      "[0.9526 0.8573 0.9341 0.9943]\r\n",
      "2023-01-05 02:53:16 [INFO]\t[EVAL] The model with the best validation mIoU (0.8813) was saved at iter 1500.\r\n",
      "2023-01-05 02:53:21 [INFO]\t[TRAIN] epoch: 69, iter: 1510/2000, loss: 0.0443, lr: 0.002825, batch_cost: 0.5845, reader_cost: 0.08602, ips: 6.8431 samples/sec | ETA 00:04:46\r\n",
      "2023-01-05 02:53:28 [INFO]\t[TRAIN] epoch: 70, iter: 1520/2000, loss: 0.0396, lr: 0.002773, batch_cost: 0.7066, reader_cost: 0.20762, ips: 5.6611 samples/sec | ETA 00:05:39\r\n",
      "2023-01-05 02:53:35 [INFO]\t[TRAIN] epoch: 70, iter: 1530/2000, loss: 0.0447, lr: 0.002721, batch_cost: 0.6861, reader_cost: 0.18780, ips: 5.8301 samples/sec | ETA 00:05:22\r\n",
      "2023-01-05 02:53:42 [INFO]\t[TRAIN] epoch: 70, iter: 1540/2000, loss: 0.0432, lr: 0.002669, batch_cost: 0.6880, reader_cost: 0.19028, ips: 5.8141 samples/sec | ETA 00:05:16\r\n",
      "2023-01-05 02:53:49 [INFO]\t[TRAIN] epoch: 71, iter: 1550/2000, loss: 0.0358, lr: 0.002617, batch_cost: 0.6873, reader_cost: 0.18884, ips: 5.8198 samples/sec | ETA 00:05:09\r\n",
      "2023-01-05 02:53:56 [INFO]\t[TRAIN] epoch: 71, iter: 1560/2000, loss: 0.0492, lr: 0.002565, batch_cost: 0.6941, reader_cost: 0.19577, ips: 5.7627 samples/sec | ETA 00:05:05\r\n",
      "2023-01-05 02:54:03 [INFO]\t[TRAIN] epoch: 72, iter: 1570/2000, loss: 0.0457, lr: 0.002512, batch_cost: 0.6976, reader_cost: 0.19953, ips: 5.7343 samples/sec | ETA 00:04:59\r\n",
      "2023-01-05 02:54:10 [INFO]\t[TRAIN] epoch: 72, iter: 1580/2000, loss: 0.0584, lr: 0.002460, batch_cost: 0.6876, reader_cost: 0.18984, ips: 5.8170 samples/sec | ETA 00:04:48\r\n",
      "2023-01-05 02:54:17 [INFO]\t[TRAIN] epoch: 73, iter: 1590/2000, loss: 0.0574, lr: 0.002407, batch_cost: 0.6969, reader_cost: 0.19903, ips: 5.7394 samples/sec | ETA 00:04:45\r\n",
      "2023-01-05 02:54:24 [INFO]\t[TRAIN] epoch: 73, iter: 1600/2000, loss: 0.0612, lr: 0.002355, batch_cost: 0.6917, reader_cost: 0.19371, ips: 5.7830 samples/sec | ETA 00:04:36\r\n",
      "2023-01-05 02:54:31 [INFO]\t[TRAIN] epoch: 74, iter: 1610/2000, loss: 0.0422, lr: 0.002302, batch_cost: 0.7042, reader_cost: 0.20775, ips: 5.6801 samples/sec | ETA 00:04:34\r\n",
      "2023-01-05 02:54:38 [INFO]\t[TRAIN] epoch: 74, iter: 1620/2000, loss: 0.0436, lr: 0.002249, batch_cost: 0.6831, reader_cost: 0.18451, ips: 5.8553 samples/sec | ETA 00:04:19\r\n",
      "2023-01-05 02:54:45 [INFO]\t[TRAIN] epoch: 75, iter: 1630/2000, loss: 0.0498, lr: 0.002195, batch_cost: 0.6925, reader_cost: 0.19462, ips: 5.7759 samples/sec | ETA 00:04:16\r\n",
      "2023-01-05 02:54:51 [INFO]\t[TRAIN] epoch: 75, iter: 1640/2000, loss: 0.0391, lr: 0.002142, batch_cost: 0.6889, reader_cost: 0.18853, ips: 5.8060 samples/sec | ETA 00:04:08\r\n",
      "2023-01-05 02:54:58 [INFO]\t[TRAIN] epoch: 75, iter: 1650/2000, loss: 0.0433, lr: 0.002089, batch_cost: 0.6879, reader_cost: 0.19092, ips: 5.8149 samples/sec | ETA 00:04:00\r\n",
      "2023-01-05 02:55:05 [INFO]\t[TRAIN] epoch: 76, iter: 1660/2000, loss: 0.0552, lr: 0.002035, batch_cost: 0.7052, reader_cost: 0.20606, ips: 5.6719 samples/sec | ETA 00:03:59\r\n",
      "2023-01-05 02:55:12 [INFO]\t[TRAIN] epoch: 76, iter: 1670/2000, loss: 0.0335, lr: 0.001981, batch_cost: 0.6916, reader_cost: 0.19418, ips: 5.7834 samples/sec | ETA 00:03:48\r\n",
      "2023-01-05 02:55:19 [INFO]\t[TRAIN] epoch: 77, iter: 1680/2000, loss: 0.0386, lr: 0.001927, batch_cost: 0.7039, reader_cost: 0.20623, ips: 5.6828 samples/sec | ETA 00:03:45\r\n",
      "2023-01-05 02:55:26 [INFO]\t[TRAIN] epoch: 77, iter: 1690/2000, loss: 0.0507, lr: 0.001873, batch_cost: 0.6808, reader_cost: 0.18244, ips: 5.8752 samples/sec | ETA 00:03:31\r\n",
      "2023-01-05 02:55:33 [INFO]\t[TRAIN] epoch: 78, iter: 1700/2000, loss: 0.0552, lr: 0.001819, batch_cost: 0.6954, reader_cost: 0.19625, ips: 5.7521 samples/sec | ETA 00:03:28\r\n",
      "2023-01-05 02:55:40 [INFO]\t[TRAIN] epoch: 78, iter: 1710/2000, loss: 0.0491, lr: 0.001764, batch_cost: 0.6766, reader_cost: 0.17728, ips: 5.9118 samples/sec | ETA 00:03:16\r\n",
      "2023-01-05 02:55:47 [INFO]\t[TRAIN] epoch: 79, iter: 1720/2000, loss: 0.0359, lr: 0.001710, batch_cost: 0.6842, reader_cost: 0.18525, ips: 5.8460 samples/sec | ETA 00:03:11\r\n",
      "2023-01-05 02:55:53 [INFO]\t[TRAIN] epoch: 79, iter: 1730/2000, loss: 0.0373, lr: 0.001655, batch_cost: 0.6816, reader_cost: 0.18351, ips: 5.8682 samples/sec | ETA 00:03:04\r\n",
      "2023-01-05 02:56:00 [INFO]\t[TRAIN] epoch: 80, iter: 1740/2000, loss: 0.0393, lr: 0.001600, batch_cost: 0.6973, reader_cost: 0.19866, ips: 5.7361 samples/sec | ETA 00:03:01\r\n",
      "2023-01-05 02:56:07 [INFO]\t[TRAIN] epoch: 80, iter: 1750/2000, loss: 0.0483, lr: 0.001544, batch_cost: 0.6923, reader_cost: 0.19476, ips: 5.7778 samples/sec | ETA 00:02:53\r\n",
      "2023-01-05 02:56:14 [INFO]\t[TRAIN] epoch: 80, iter: 1760/2000, loss: 0.0431, lr: 0.001489, batch_cost: 0.6830, reader_cost: 0.18612, ips: 5.8569 samples/sec | ETA 00:02:43\r\n",
      "2023-01-05 02:56:21 [INFO]\t[TRAIN] epoch: 81, iter: 1770/2000, loss: 0.0428, lr: 0.001433, batch_cost: 0.6959, reader_cost: 0.19690, ips: 5.7479 samples/sec | ETA 00:02:40\r\n",
      "2023-01-05 02:56:28 [INFO]\t[TRAIN] epoch: 81, iter: 1780/2000, loss: 0.0415, lr: 0.001377, batch_cost: 0.6805, reader_cost: 0.18258, ips: 5.8784 samples/sec | ETA 00:02:29\r\n",
      "2023-01-05 02:56:35 [INFO]\t[TRAIN] epoch: 82, iter: 1790/2000, loss: 0.0453, lr: 0.001321, batch_cost: 0.7056, reader_cost: 0.20682, ips: 5.6689 samples/sec | ETA 00:02:28\r\n",
      "2023-01-05 02:56:42 [INFO]\t[TRAIN] epoch: 82, iter: 1800/2000, loss: 0.0371, lr: 0.001265, batch_cost: 0.6893, reader_cost: 0.19058, ips: 5.8033 samples/sec | ETA 00:02:17\r\n",
      "2023-01-05 02:56:49 [INFO]\t[TRAIN] epoch: 83, iter: 1810/2000, loss: 0.0442, lr: 0.001208, batch_cost: 0.6939, reader_cost: 0.19441, ips: 5.7647 samples/sec | ETA 00:02:11\r\n",
      "2023-01-05 02:56:56 [INFO]\t[TRAIN] epoch: 83, iter: 1820/2000, loss: 0.0421, lr: 0.001151, batch_cost: 0.6747, reader_cost: 0.17967, ips: 5.9289 samples/sec | ETA 00:02:01\r\n",
      "2023-01-05 02:57:03 [INFO]\t[TRAIN] epoch: 84, iter: 1830/2000, loss: 0.0510, lr: 0.001093, batch_cost: 0.6953, reader_cost: 0.19606, ips: 5.7528 samples/sec | ETA 00:01:58\r\n",
      "2023-01-05 02:57:09 [INFO]\t[TRAIN] epoch: 84, iter: 1840/2000, loss: 0.0473, lr: 0.001036, batch_cost: 0.6922, reader_cost: 0.19396, ips: 5.7784 samples/sec | ETA 00:01:50\r\n",
      "2023-01-05 02:57:16 [INFO]\t[TRAIN] epoch: 85, iter: 1850/2000, loss: 0.0424, lr: 0.000978, batch_cost: 0.6884, reader_cost: 0.19025, ips: 5.8108 samples/sec | ETA 00:01:43\r\n",
      "2023-01-05 02:57:23 [INFO]\t[TRAIN] epoch: 85, iter: 1860/2000, loss: 0.0397, lr: 0.000919, batch_cost: 0.6747, reader_cost: 0.17798, ips: 5.9282 samples/sec | ETA 00:01:34\r\n",
      "2023-01-05 02:57:30 [INFO]\t[TRAIN] epoch: 85, iter: 1870/2000, loss: 0.0437, lr: 0.000860, batch_cost: 0.6820, reader_cost: 0.18364, ips: 5.8650 samples/sec | ETA 00:01:28\r\n",
      "2023-01-05 02:57:37 [INFO]\t[TRAIN] epoch: 86, iter: 1880/2000, loss: 0.0326, lr: 0.000801, batch_cost: 0.6930, reader_cost: 0.19373, ips: 5.7719 samples/sec | ETA 00:01:23\r\n",
      "2023-01-05 02:57:44 [INFO]\t[TRAIN] epoch: 86, iter: 1890/2000, loss: 0.0440, lr: 0.000741, batch_cost: 0.6750, reader_cost: 0.17680, ips: 5.9262 samples/sec | ETA 00:01:14\r\n",
      "2023-01-05 02:57:51 [INFO]\t[TRAIN] epoch: 87, iter: 1900/2000, loss: 0.0333, lr: 0.000681, batch_cost: 0.6915, reader_cost: 0.19413, ips: 5.7849 samples/sec | ETA 00:01:09\r\n",
      "2023-01-05 02:57:57 [INFO]\t[TRAIN] epoch: 87, iter: 1910/2000, loss: 0.0421, lr: 0.000620, batch_cost: 0.6923, reader_cost: 0.19241, ips: 5.7782 samples/sec | ETA 00:01:02\r\n",
      "2023-01-05 02:58:04 [INFO]\t[TRAIN] epoch: 88, iter: 1920/2000, loss: 0.0388, lr: 0.000558, batch_cost: 0.6949, reader_cost: 0.19767, ips: 5.7564 samples/sec | ETA 00:00:55\r\n",
      "2023-01-05 02:58:11 [INFO]\t[TRAIN] epoch: 88, iter: 1930/2000, loss: 0.0408, lr: 0.000496, batch_cost: 0.6853, reader_cost: 0.18748, ips: 5.8367 samples/sec | ETA 00:00:47\r\n",
      "2023-01-05 02:58:18 [INFO]\t[TRAIN] epoch: 89, iter: 1940/2000, loss: 0.0374, lr: 0.000432, batch_cost: 0.6908, reader_cost: 0.19316, ips: 5.7905 samples/sec | ETA 00:00:41\r\n",
      "2023-01-05 02:58:25 [INFO]\t[TRAIN] epoch: 89, iter: 1950/2000, loss: 0.0402, lr: 0.000368, batch_cost: 0.6719, reader_cost: 0.17565, ips: 5.9532 samples/sec | ETA 00:00:33\r\n",
      "2023-01-05 02:58:32 [INFO]\t[TRAIN] epoch: 90, iter: 1960/2000, loss: 0.0429, lr: 0.000302, batch_cost: 0.6866, reader_cost: 0.18965, ips: 5.8261 samples/sec | ETA 00:00:27\r\n",
      "2023-01-05 02:58:39 [INFO]\t[TRAIN] epoch: 90, iter: 1970/2000, loss: 0.0618, lr: 0.000235, batch_cost: 0.6926, reader_cost: 0.19569, ips: 5.7752 samples/sec | ETA 00:00:20\r\n",
      "2023-01-05 02:58:46 [INFO]\t[TRAIN] epoch: 90, iter: 1980/2000, loss: 0.0381, lr: 0.000166, batch_cost: 0.6865, reader_cost: 0.18934, ips: 5.8264 samples/sec | ETA 00:00:13\r\n",
      "2023-01-05 02:58:53 [INFO]\t[TRAIN] epoch: 91, iter: 1990/2000, loss: 0.0422, lr: 0.000093, batch_cost: 0.6951, reader_cost: 0.19642, ips: 5.7543 samples/sec | ETA 00:00:06\r\n",
      "2023-01-05 02:58:59 [INFO]\t[TRAIN] epoch: 91, iter: 2000/2000, loss: 0.0397, lr: 0.000011, batch_cost: 0.6756, reader_cost: 0.17854, ips: 5.9208 samples/sec | ETA 00:00:00\r\n",
      "2023-01-05 02:58:59 [INFO]\tStart evaluating (total_samples: 10, total_iters: 10)...\r\n",
      "10/10 [==============================] - 1s 78ms/step - batch_cost: 0.0783 - reader cost: 0.0425\r\n",
      "2023-01-05 02:59:00 [INFO]\t[EVAL] #Images: 10 mIoU: 0.8826 Acc: 0.9881 Kappa: 0.9295 Dice: 0.9362\r\n",
      "2023-01-05 02:59:00 [INFO]\t[EVAL] Class IoU: \r\n",
      "[0.8843 0.7958 0.861  0.9893]\r\n",
      "2023-01-05 02:59:00 [INFO]\t[EVAL] Class Precision: \r\n",
      "[0.9304 0.9062 0.9229 0.9946]\r\n",
      "2023-01-05 02:59:00 [INFO]\t[EVAL] Class Recall: \r\n",
      "[0.947  0.8672 0.9278 0.9947]\r\n",
      "2023-01-05 02:59:01 [INFO]\t[EVAL] The model with the best validation mIoU (0.8826) was saved at iter 2000.\r\n",
      "<class 'paddle.nn.layer.conv.Conv2D'>'s flops has been counted\r\n",
      "<class 'paddle.nn.layer.norm.BatchNorm2D'>'s flops has been counted\r\n",
      "<class 'paddle.nn.layer.activation.ReLU'>'s flops has been counted\r\n",
      "Cannot find suitable count function for <class 'paddle.nn.layer.pooling.MaxPool2D'>. Treat it as zero FLOPs.\r\n",
      "Total Flops: 249212960768     Total Params: 13405508\r\n"
     ]
    }
   ],
   "source": [
    "%cd /home/aistudio/PaddleSeg/\n",
    "\n",
    "!export CUDA_VISIBLE_DEVICES=0 # 设置1张可用的卡\n",
    "\n",
    "!python train.py \\\n",
    "       --config configs/unet/unet_cityscapes_1024x512_160k.yml \\\n",
    "       --save_interval 500 \\\n",
    "       --do_eval \\\n",
    "       --use_vdl \\\n",
    "       --save_dir output\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 四.评估"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-04T18:59:10.932311Z",
     "iopub.status.busy": "2023-01-04T18:59:10.931170Z",
     "iopub.status.idle": "2023-01-04T18:59:17.148615Z",
     "shell.execute_reply": "2023-01-04T18:59:17.147575Z",
     "shell.execute_reply.started": "2023-01-04T18:59:10.932237Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2023-01-05 02:59:13 [INFO]\t\r\n",
      "---------------Config Information---------------\r\n",
      "batch_size: 4\r\n",
      "iters: 2000\r\n",
      "loss:\r\n",
      "  coef:\r\n",
      "  - 1\r\n",
      "  types:\r\n",
      "  - type: CrossEntropyLoss\r\n",
      "lr_scheduler:\r\n",
      "  end_lr: 0\r\n",
      "  learning_rate: 0.01\r\n",
      "  power: 0.9\r\n",
      "  type: PolynomialDecay\r\n",
      "model:\r\n",
      "  num_classes: 4\r\n",
      "  pretrained: null\r\n",
      "  type: UNet\r\n",
      "  use_deconv: false\r\n",
      "optimizer:\r\n",
      "  momentum: 0.9\r\n",
      "  type: sgd\r\n",
      "  weight_decay: 4.0e-05\r\n",
      "train_dataset:\r\n",
      "  dataset_root: data/mine\r\n",
      "  mode: train\r\n",
      "  num_classes: 4\r\n",
      "  train_path: data/mine/train.txt\r\n",
      "  transforms:\r\n",
      "  - max_scale_factor: 2.0\r\n",
      "    min_scale_factor: 0.5\r\n",
      "    scale_step_size: 0.25\r\n",
      "    type: ResizeStepScaling\r\n",
      "  - crop_size:\r\n",
      "    - 1024\r\n",
      "    - 512\r\n",
      "    type: RandomPaddingCrop\r\n",
      "  - type: RandomHorizontalFlip\r\n",
      "  - brightness_range: 0.4\r\n",
      "    contrast_range: 0.4\r\n",
      "    saturation_range: 0.4\r\n",
      "    type: RandomDistort\r\n",
      "  - type: Normalize\r\n",
      "  type: Dataset\r\n",
      "val_dataset:\r\n",
      "  dataset_root: data/mine\r\n",
      "  mode: val\r\n",
      "  num_classes: 4\r\n",
      "  transforms:\r\n",
      "  - type: Normalize\r\n",
      "  type: Dataset\r\n",
      "  val_path: data/mine/valid.txt\r\n",
      "------------------------------------------------\r\n",
      "W0105 02:59:13.092689 21272 gpu_resources.cc:61] Please NOTE: device: 0, GPU Compute Capability: 7.0, Driver API Version: 11.2, Runtime API Version: 11.2\r\n",
      "W0105 02:59:13.092732 21272 gpu_resources.cc:91] device: 0, cuDNN Version: 8.2.\r\n",
      "2023-01-05 02:59:14 [INFO]\tLoading pretrained model from output/best_model/model.pdparams\r\n",
      "2023-01-05 02:59:14 [INFO]\tThere are 112/112 variables loaded into UNet.\r\n",
      "2023-01-05 02:59:14 [INFO]\tLoaded trained params of model successfully\r\n",
      "2023-01-05 02:59:14 [INFO]\tStart evaluating (total_samples: 10, total_iters: 10)...\r\n",
      "10/10 [==============================] - 2s 189ms/step - batch_cost: 0.1888 - reader cost: 0.0207\r\n",
      "2023-01-05 02:59:16 [INFO]\t[EVAL] #Images: 10 mIoU: 0.8826 Acc: 0.9881 Kappa: 0.9295 Dice: 0.9362\r\n",
      "2023-01-05 02:59:16 [INFO]\t[EVAL] Class IoU: \r\n",
      "[0.8843 0.7958 0.861  0.9893]\r\n",
      "2023-01-05 02:59:16 [INFO]\t[EVAL] Class Precision: \r\n",
      "[0.9304 0.9062 0.9229 0.9946]\r\n",
      "2023-01-05 02:59:16 [INFO]\t[EVAL] Class Recall: \r\n",
      "[0.947  0.8672 0.9278 0.9947]\r\n"
     ]
    }
   ],
   "source": [
    "!python val.py \\\n",
    "       --config configs/unet/unet_cityscapes_1024x512_160k.yml \\\n",
    "       --model_path output/best_model/model.pdparams"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 五.A榜图片"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**需要注意的是，我们在前面数据增强使用了windowing方案，则推理时必须做同样的处理**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-04T18:59:25.125056Z",
     "iopub.status.busy": "2023-01-04T18:59:25.124136Z",
     "iopub.status.idle": "2023-01-04T18:59:28.428942Z",
     "shell.execute_reply": "2023-01-04T18:59:28.427854Z",
     "shell.execute_reply.started": "2023-01-04T18:59:25.125020Z"
    },
    "scrolled": true,
    "tags": []
   },
   "outputs": [],
   "source": [
    "!mkdir /home/aistudio/PaddleSeg/data/mine/test\n",
    "testWin = windowImg('test')\n",
    "testWin.run()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-04T18:59:36.299258Z",
     "iopub.status.busy": "2023-01-04T18:59:36.298557Z",
     "iopub.status.idle": "2023-01-04T18:59:56.570836Z",
     "shell.execute_reply": "2023-01-04T18:59:56.569780Z",
     "shell.execute_reply.started": "2023-01-04T18:59:36.299225Z"
    },
    "scrolled": true,
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2023-01-05 02:59:38 [INFO]\t\r\n",
      "---------------Config Information---------------\r\n",
      "batch_size: 4\r\n",
      "iters: 2000\r\n",
      "loss:\r\n",
      "  coef:\r\n",
      "  - 1\r\n",
      "  types:\r\n",
      "  - type: CrossEntropyLoss\r\n",
      "lr_scheduler:\r\n",
      "  end_lr: 0\r\n",
      "  learning_rate: 0.01\r\n",
      "  power: 0.9\r\n",
      "  type: PolynomialDecay\r\n",
      "model:\r\n",
      "  num_classes: 4\r\n",
      "  pretrained: null\r\n",
      "  type: UNet\r\n",
      "  use_deconv: false\r\n",
      "optimizer:\r\n",
      "  momentum: 0.9\r\n",
      "  type: sgd\r\n",
      "  weight_decay: 4.0e-05\r\n",
      "train_dataset:\r\n",
      "  dataset_root: data/mine\r\n",
      "  mode: train\r\n",
      "  num_classes: 4\r\n",
      "  train_path: data/mine/train.txt\r\n",
      "  transforms:\r\n",
      "  - max_scale_factor: 2.0\r\n",
      "    min_scale_factor: 0.5\r\n",
      "    scale_step_size: 0.25\r\n",
      "    type: ResizeStepScaling\r\n",
      "  - crop_size:\r\n",
      "    - 1024\r\n",
      "    - 512\r\n",
      "    type: RandomPaddingCrop\r\n",
      "  - type: RandomHorizontalFlip\r\n",
      "  - brightness_range: 0.4\r\n",
      "    contrast_range: 0.4\r\n",
      "    saturation_range: 0.4\r\n",
      "    type: RandomDistort\r\n",
      "  - type: Normalize\r\n",
      "  type: Dataset\r\n",
      "val_dataset:\r\n",
      "  dataset_root: data/mine\r\n",
      "  mode: val\r\n",
      "  num_classes: 4\r\n",
      "  transforms:\r\n",
      "  - type: Normalize\r\n",
      "  type: Dataset\r\n",
      "  val_path: data/mine/valid.txt\r\n",
      "------------------------------------------------\r\n",
      "W0105 02:59:38.265364 21387 gpu_resources.cc:61] Please NOTE: device: 0, GPU Compute Capability: 7.0, Driver API Version: 11.2, Runtime API Version: 11.2\r\n",
      "W0105 02:59:38.265408 21387 gpu_resources.cc:91] device: 0, cuDNN Version: 8.2.\r\n",
      "2023-01-05 02:59:39 [INFO]\tNumber of predict images = 100\r\n",
      "2023-01-05 02:59:39 [INFO]\tLoading pretrained model from output/best_model/model.pdparams\r\n",
      "2023-01-05 02:59:39 [INFO]\tThere are 112/112 variables loaded into UNet.\r\n",
      "2023-01-05 02:59:39 [INFO]\tStart to predict...\r\n",
      "100/100 [==============================] - 16s 162ms/step\r\n"
     ]
    }
   ],
   "source": [
    "!python predict.py \\\n",
    "       --config      configs/unet/unet_cityscapes_1024x512_160k.yml \\\n",
    "       --model_path output/best_model/model.pdparams \\\n",
    "       --image_path data/mine/test \\\n",
    "       --save_dir output/result \\\n",
    "       --custom_color 0 0 0 80 80 80 160 160 160 255 255 255\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 六.提交比赛结果\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-04T19:02:27.334280Z",
     "iopub.status.busy": "2023-01-04T19:02:27.333461Z",
     "iopub.status.idle": "2023-01-04T19:02:28.033640Z",
     "shell.execute_reply": "2023-01-04T19:02:28.032655Z",
     "shell.execute_reply.started": "2023-01-04T19:02:27.334244Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  adding: home/aistudio/PaddleSeg/output/result/ (stored 0%)\r\n"
     ]
    }
   ],
   "source": [
    "!rm -rf /home/aistudio/PaddleSeg/output/result/added_prediction\n",
    "!mv /home/aistudio/PaddleSeg/output/result/pseudo_color_prediction Layer_Segmentations\n",
    "!zip -r result.zip  /home/aistudio/PaddleSeg/output/result/\n",
    "#最后提交文件即可。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 七.总结\n",
    "\n",
    "分数其实并没有超过前几个月的选手，主要是多了一个数据增强的方案。OCT是一个很热门的技术，作者在做脑部OCT的一些工作，主要是数据处理以及器械。OCT的图像采集是一个很重要的工作，本文的应用场景是眼，其实不仅仅是眼。脑部血管也可以用。最后希望自己能早日获取PPDE。\n",
    "\n",
    "参考链接：\n",
    "\n",
    "[飞桨学习赛：眼底OCT层分割（GOALS挑战赛任务一）第一名方案](https://aistudio.baidu.com/aistudio/projectdetail/4609530?channelType=0&channel=0)\n",
    "\n",
    "[分割网络](https://aistudio.baidu.com/aistudio/projectdetail/2148971?channelType=0&channel=0)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "py35-paddle1.2.0"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}