U-Net_MNIST.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "35787630-2223-464b-bbc9-bd1cf6f2ef25",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.utils import plot_model\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "from tensorflow.keras.layers import Input, Add, Dense, Activation, ZeroPadding2D, BatchNormalization,\\\n",
    "     Flatten, Conv2D, MaxPooling2D, UpSampling2D, Dropout, concatenate\n",
    "from tensorflow.keras import layers\n",
    "from tensorflow.keras.initializers import glorot_uniform\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras.datasets import mnist\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.callbacks import *\n",
    "# from tensorflow.python.framework.ops import disable_eager_execution\n",
    "# disable_eager_execution()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "0e591e21",
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "6af4df94",
   "metadata": {},
   "outputs": [],
   "source": [
    "(train_images, train_labels), (test_images, test_labels) = mnist.load_data()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "0c4130d8",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_images = train_images/255.0\n",
    "test_images = test_images/255.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "81857741",
   "metadata": {},
   "outputs": [],
   "source": [
    "w, h = 28, 28\n",
    "train_images = train_images.reshape(train_images.shape[0], w, h, 1)\n",
    "test_images = test_images.reshape(test_images.shape[0], w, h, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "04bff7fe",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60000, 28, 28, 1)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_images.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "74f22a7a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Default GPU Device:/device:GPU:0\n"
     ]
    }
   ],
   "source": [
    "if tf.test.gpu_device_name(): \n",
    "\n",
    "    print('Default GPU Device:{}'.format(tf.test.gpu_device_name()))\n",
    "\n",
    "else:\n",
    "    print(\"Please install GPU version of TF\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "b8935258",
   "metadata": {},
   "outputs": [],
   "source": [
    "classess = ['0', '1', '2', '3', '4', '5', '6', '7', '8', '9']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "2d7f70f7",
   "metadata": {},
   "outputs": [],
   "source": [
    "def Structure(X):\n",
    "    conv1 = Conv2D(16, 3, activation = 'relu', padding = 'same', kernel_initializer = glorot_uniform(seed = 0))(X)\n",
    "    conv1 = Conv2D(16, 3, activation='relu', padding='same')(conv1)\n",
    "    pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)\n",
    "    pool1 = Dropout(0.2)(pool1)\n",
    "    conv2 = Conv2D(32, 3, activation='relu', padding='same')(pool1)\n",
    "    conv2 = Conv2D(32, 3, activation='relu', padding='same')(conv2)\n",
    "    pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)\n",
    "    pool2 = Dropout(0.3)(pool2)\n",
    "    # conv3 = Conv2D(64, 2, activation='relu', padding='same')(pool2)\n",
    "    # conv3 = Conv2D(64, 2, activation='relu', padding='same')(conv3)\n",
    "    # conv3 = Conv2D(64, 2, activation='relu', padding='same')(conv3)\n",
    "    # pool3 = MaxPooling2D(pool_size=(2, 2))(conv3)\n",
    "    # X = Conv2D(256, 3, activation='relu', padding='same')(X)\n",
    "    # X = Conv2D(256, 3, activation='relu', padding='same')(X)\n",
    "    # X = Conv2D(256, 3, activation='relu', padding='same')(X)\n",
    "    # X = MaxPooling2D(pool_size=(2, 2))(X)\n",
    "    # X = Conv2D(128, 3, activation='relu', padding='same')(X)\n",
    "    # X = Conv2D(128, 3, activation='relu', padding='same')(X)\n",
    "    # X = Conv2D(128, 3, activation='relu', padding='same')(X)\n",
    "    # X = MaxPooling2D(pool_size=(2, 2))(X)\n",
    "    # X = UpSampling2D(size=(2, 2))(X)\n",
    "    # X = Conv2D(128, 3, activation='relu', padding='same')(X)\n",
    "    # X = Conv2D(128, 3, activation='relu', padding='same')(X)\n",
    "    # X = Conv2D(128, 3, activation='relu', padding='same')(X)   \n",
    "    # X = UpSampling2D(size=(2, 2))(X)\n",
    "    # X = Conv2D(256, 3, activation='relu', padding='same')(X)\n",
    "    # X = Conv2D(256, 3, activation='relu', padding='same')(X)\n",
    "    # X = Conv2D(256, 3, activation='relu', padding='same')(X)   \n",
    "    # up1 = UpSampling2D(size=(2, 2))(pool3)\n",
    "    # up1 = concatenate([conv3, up1], axis=-1)\n",
    "    # conv4 = Conv2D(64, 3, activation='relu', padding='same')(up1)\n",
    "    # conv4 = Conv2D(64, 3, activation='relu', padding='same')(conv4)\n",
    "    # conv4 = Conv2D(64, 3, activation='relu', padding='same')(conv4)   \n",
    "    # conv4 = Dropout(0.3)(conv4)\n",
    "    up2 = UpSampling2D(size=(2, 2))(pool2)\n",
    "    up2 = concatenate([conv2, up2], axis=-1)\n",
    "    conv5 = Conv2D(32, 3, activation='relu', padding='same')(up2)\n",
    "    conv5 = Conv2D(32, 3, activation='relu', padding='same')(conv5)\n",
    "    conv5 = Dropout(0.2)(conv5)\n",
    "    up3 = UpSampling2D(size=(2, 2))(conv5)\n",
    "    up3 = concatenate([conv1, up3], axis=-1)\n",
    "    conv6 = Conv2D(16, 3, activation='relu', padding='same')(up3)\n",
    "    conv6 = Conv2D(16, 3, activation='relu', padding='same')(conv6)\n",
    "    return conv6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "4fccd0e9",
   "metadata": {},
   "outputs": [],
   "source": [
    "def UNET(input_shape = (w, h, 1), classes = classess):\n",
    "    X_input = Input(shape = input_shape)\n",
    "    X = Structure(X_input)\n",
    "    X = Conv2D(len(classes), 1, activation = 'softmax')(X)\n",
    "    X = Flatten()(X)\n",
    "    X = Dense(len(classess), activation='softmax')(X)\n",
    "    model = Model(inputs = X_input, outputs = X, name = 'U-net')\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "e99b66e6",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Clear unnecessary memory\n",
    "from tensorflow.keras import backend as K\n",
    "import tensorflow as tf\n",
    "K.clear_session()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "2a78b2c8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"U-net\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            [(None, 28, 28, 1)]  0                                            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d (Conv2D)                 (None, 28, 28, 16)   160         input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_1 (Conv2D)               (None, 28, 28, 16)   2320        conv2d[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d (MaxPooling2D)    (None, 14, 14, 16)   0           conv2d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout (Dropout)               (None, 14, 14, 16)   0           max_pooling2d[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_2 (Conv2D)               (None, 14, 14, 32)   4640        dropout[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_3 (Conv2D)               (None, 14, 14, 32)   9248        conv2d_2[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2D)  (None, 7, 7, 32)     0           conv2d_3[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout_1 (Dropout)             (None, 7, 7, 32)     0           max_pooling2d_1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "up_sampling2d (UpSampling2D)    (None, 14, 14, 32)   0           dropout_1[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate (Concatenate)       (None, 14, 14, 64)   0           conv2d_3[0][0]                   \n",
      "                                                                 up_sampling2d[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_4 (Conv2D)               (None, 14, 14, 32)   18464       concatenate[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_5 (Conv2D)               (None, 14, 14, 32)   9248        conv2d_4[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout_2 (Dropout)             (None, 14, 14, 32)   0           conv2d_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "up_sampling2d_1 (UpSampling2D)  (None, 28, 28, 32)   0           dropout_2[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_1 (Concatenate)     (None, 28, 28, 48)   0           conv2d_1[0][0]                   \n",
      "                                                                 up_sampling2d_1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_6 (Conv2D)               (None, 28, 28, 16)   6928        concatenate_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_7 (Conv2D)               (None, 28, 28, 16)   2320        conv2d_6[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_8 (Conv2D)               (None, 28, 28, 10)   170         conv2d_7[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "flatten (Flatten)               (None, 7840)         0           conv2d_8[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dense (Dense)                   (None, 10)           78410       flatten[0][0]                    \n",
      "==================================================================================================\n",
      "Total params: 131,908\n",
      "Trainable params: 131,908\n",
      "Non-trainable params: 0\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = UNET(input_shape = (w, h, 1), classes = classess)\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "6a436d92",
   "metadata": {},
   "outputs": [],
   "source": [
    "# import os\n",
    "# import datetime\n",
    "\n",
    "# logdir = os.path.join(\"logs\", datetime.datetime.now().strftime(\"%Y%m%d-%H%M%S\"))\n",
    "# tensorboard_callback = tf.keras.callbacks.TensorBoard(logdir, histogram_freq=1)\n",
    "callbacks = [\n",
    "    EarlyStopping(patience=10, verbose=1),\n",
    "    ReduceLROnPlateau(factor=0.1, patience=5, min_lr=0.00001, verbose=1),\n",
    "    ModelCheckpoint('model-Unet-mnist.h5', verbose=1, save_best_only=True, save_weights_only=True),\n",
    "    CSVLogger(\"dataUnetmnist.csv\"),\n",
    "    TensorBoard(log_dir='./logs-Unet-MNIST', histogram_freq=1)\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "eaf489db",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compile the model\n",
    "model.compile(optimizer='adam',\n",
    "              loss= 'sparse_categorical_crossentropy',\n",
    "              metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "45ebc787",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "200/200 [==============================] - 15s 23ms/step - loss: 0.4985 - accuracy: 0.8405 - val_loss: 0.0991 - val_accuracy: 0.9719\n",
      "\n",
      "Epoch 00001: val_loss improved from inf to 0.09910, saving model to model-Unet-mnist.h5\n",
      "Epoch 2/100\n",
      "200/200 [==============================] - 4s 18ms/step - loss: 0.0958 - accuracy: 0.9712 - val_loss: 0.0643 - val_accuracy: 0.9817\n",
      "\n",
      "Epoch 00002: val_loss improved from 0.09910 to 0.06431, saving model to model-Unet-mnist.h5\n",
      "Epoch 3/100\n",
      "200/200 [==============================] - 4s 18ms/step - loss: 0.0706 - accuracy: 0.9777 - val_loss: 0.0525 - val_accuracy: 0.9856\n",
      "\n",
      "Epoch 00003: val_loss improved from 0.06431 to 0.05246, saving model to model-Unet-mnist.h5\n",
      "Epoch 4/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0563 - accuracy: 0.9825 - val_loss: 0.0512 - val_accuracy: 0.9851\n",
      "\n",
      "Epoch 00004: val_loss improved from 0.05246 to 0.05117, saving model to model-Unet-mnist.h5\n",
      "Epoch 5/100\n",
      "200/200 [==============================] - 3s 15ms/step - loss: 0.0489 - accuracy: 0.9844 - val_loss: 0.0466 - val_accuracy: 0.9871\n",
      "\n",
      "Epoch 00005: val_loss improved from 0.05117 to 0.04661, saving model to model-Unet-mnist.h5\n",
      "Epoch 6/100\n",
      "200/200 [==============================] - 3s 15ms/step - loss: 0.0431 - accuracy: 0.9871 - val_loss: 0.0459 - val_accuracy: 0.9866\n",
      "\n",
      "Epoch 00006: val_loss improved from 0.04661 to 0.04586, saving model to model-Unet-mnist.h5\n",
      "Epoch 7/100\n",
      "200/200 [==============================] - 3s 16ms/step - loss: 0.0379 - accuracy: 0.9883 - val_loss: 0.0392 - val_accuracy: 0.9891\n",
      "\n",
      "Epoch 00007: val_loss improved from 0.04586 to 0.03915, saving model to model-Unet-mnist.h5\n",
      "Epoch 8/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0349 - accuracy: 0.9895 - val_loss: 0.0467 - val_accuracy: 0.9877\n",
      "\n",
      "Epoch 00008: val_loss did not improve from 0.03915\n",
      "Epoch 9/100\n",
      "200/200 [==============================] - 3s 16ms/step - loss: 0.0335 - accuracy: 0.9890 - val_loss: 0.0383 - val_accuracy: 0.9890\n",
      "\n",
      "Epoch 00009: val_loss improved from 0.03915 to 0.03834, saving model to model-Unet-mnist.h5\n",
      "Epoch 10/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0280 - accuracy: 0.9912 - val_loss: 0.0387 - val_accuracy: 0.9893\n",
      "\n",
      "Epoch 00010: val_loss did not improve from 0.03834\n",
      "Epoch 11/100\n",
      "200/200 [==============================] - 3s 15ms/step - loss: 0.0279 - accuracy: 0.9907 - val_loss: 0.0359 - val_accuracy: 0.9898\n",
      "\n",
      "Epoch 00011: val_loss improved from 0.03834 to 0.03586, saving model to model-Unet-mnist.h5\n",
      "Epoch 12/100\n",
      "200/200 [==============================] - 3s 15ms/step - loss: 0.0249 - accuracy: 0.9919 - val_loss: 0.0341 - val_accuracy: 0.9900\n",
      "\n",
      "Epoch 00012: val_loss improved from 0.03586 to 0.03414, saving model to model-Unet-mnist.h5\n",
      "Epoch 13/100\n",
      "200/200 [==============================] - 3s 15ms/step - loss: 0.0206 - accuracy: 0.9933 - val_loss: 0.0361 - val_accuracy: 0.9894\n",
      "\n",
      "Epoch 00013: val_loss did not improve from 0.03414\n",
      "Epoch 14/100\n",
      "200/200 [==============================] - 3s 15ms/step - loss: 0.0227 - accuracy: 0.9925 - val_loss: 0.0374 - val_accuracy: 0.9901\n",
      "\n",
      "Epoch 00014: val_loss did not improve from 0.03414\n",
      "Epoch 15/100\n",
      "200/200 [==============================] - 3s 15ms/step - loss: 0.0208 - accuracy: 0.9932 - val_loss: 0.0384 - val_accuracy: 0.9891\n",
      "\n",
      "Epoch 00015: val_loss did not improve from 0.03414\n",
      "Epoch 16/100\n",
      "200/200 [==============================] - 3s 15ms/step - loss: 0.0198 - accuracy: 0.9935 - val_loss: 0.0398 - val_accuracy: 0.9884\n",
      "\n",
      "Epoch 00016: val_loss did not improve from 0.03414\n",
      "Epoch 17/100\n",
      "200/200 [==============================] - 3s 15ms/step - loss: 0.0178 - accuracy: 0.9944 - val_loss: 0.0322 - val_accuracy: 0.9903\n",
      "\n",
      "Epoch 00017: val_loss improved from 0.03414 to 0.03224, saving model to model-Unet-mnist.h5\n",
      "Epoch 18/100\n",
      "200/200 [==============================] - 3s 16ms/step - loss: 0.0155 - accuracy: 0.9948 - val_loss: 0.0350 - val_accuracy: 0.9906\n",
      "\n",
      "Epoch 00018: val_loss did not improve from 0.03224\n",
      "Epoch 19/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0143 - accuracy: 0.9948 - val_loss: 0.0372 - val_accuracy: 0.9898\n",
      "\n",
      "Epoch 00019: val_loss did not improve from 0.03224\n",
      "Epoch 20/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0142 - accuracy: 0.9954 - val_loss: 0.0378 - val_accuracy: 0.9898\n",
      "\n",
      "Epoch 00020: val_loss did not improve from 0.03224\n",
      "Epoch 21/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0130 - accuracy: 0.9957 - val_loss: 0.0396 - val_accuracy: 0.9888\n",
      "\n",
      "Epoch 00021: val_loss did not improve from 0.03224\n",
      "Epoch 22/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0137 - accuracy: 0.9955 - val_loss: 0.0355 - val_accuracy: 0.9900\n",
      "\n",
      "Epoch 00022: ReduceLROnPlateau reducing learning rate to 0.00010000000474974513.\n",
      "\n",
      "Epoch 00022: val_loss did not improve from 0.03224\n",
      "Epoch 23/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0075 - accuracy: 0.9976 - val_loss: 0.0313 - val_accuracy: 0.9922\n",
      "\n",
      "Epoch 00023: val_loss improved from 0.03224 to 0.03135, saving model to model-Unet-mnist.h5\n",
      "Epoch 24/100\n",
      "200/200 [==============================] - 3s 16ms/step - loss: 0.0054 - accuracy: 0.9987 - val_loss: 0.0307 - val_accuracy: 0.9923\n",
      "\n",
      "Epoch 00024: val_loss improved from 0.03135 to 0.03071, saving model to model-Unet-mnist.h5\n",
      "Epoch 25/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0049 - accuracy: 0.9988 - val_loss: 0.0310 - val_accuracy: 0.9923\n",
      "\n",
      "Epoch 00025: val_loss did not improve from 0.03071\n",
      "Epoch 26/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0047 - accuracy: 0.9989 - val_loss: 0.0298 - val_accuracy: 0.9927\n",
      "\n",
      "Epoch 00026: val_loss improved from 0.03071 to 0.02978, saving model to model-Unet-mnist.h5\n",
      "Epoch 27/100\n",
      "200/200 [==============================] - 3s 16ms/step - loss: 0.0042 - accuracy: 0.9990 - val_loss: 0.0300 - val_accuracy: 0.9927\n",
      "\n",
      "Epoch 00027: val_loss did not improve from 0.02978\n",
      "Epoch 28/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0045 - accuracy: 0.9990 - val_loss: 0.0301 - val_accuracy: 0.9924\n",
      "\n",
      "Epoch 00028: val_loss did not improve from 0.02978\n",
      "Epoch 29/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0042 - accuracy: 0.9992 - val_loss: 0.0312 - val_accuracy: 0.9921\n",
      "\n",
      "Epoch 00029: val_loss did not improve from 0.02978\n",
      "Epoch 30/100\n",
      "200/200 [==============================] - 3s 16ms/step - loss: 0.0038 - accuracy: 0.9991 - val_loss: 0.0317 - val_accuracy: 0.9926\n",
      "\n",
      "Epoch 00030: val_loss did not improve from 0.02978\n",
      "Epoch 31/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0037 - accuracy: 0.9993 - val_loss: 0.0309 - val_accuracy: 0.9923\n",
      "\n",
      "Epoch 00031: ReduceLROnPlateau reducing learning rate to 1.0000000474974514e-05.\n",
      "\n",
      "Epoch 00031: val_loss did not improve from 0.02978\n",
      "Epoch 32/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0035 - accuracy: 0.9995 - val_loss: 0.0310 - val_accuracy: 0.9923\n",
      "\n",
      "Epoch 00032: val_loss did not improve from 0.02978\n",
      "Epoch 33/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0034 - accuracy: 0.9994 - val_loss: 0.0314 - val_accuracy: 0.9923\n",
      "\n",
      "Epoch 00033: val_loss did not improve from 0.02978\n",
      "Epoch 34/100\n",
      "200/200 [==============================] - 3s 16ms/step - loss: 0.0037 - accuracy: 0.9991 - val_loss: 0.0313 - val_accuracy: 0.9923\n",
      "\n",
      "Epoch 00034: val_loss did not improve from 0.02978\n",
      "Epoch 35/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0033 - accuracy: 0.9995 - val_loss: 0.0312 - val_accuracy: 0.9922\n",
      "\n",
      "Epoch 00035: val_loss did not improve from 0.02978\n",
      "Epoch 36/100\n",
      "200/200 [==============================] - 3s 17ms/step - loss: 0.0033 - accuracy: 0.9994 - val_loss: 0.0311 - val_accuracy: 0.9926\n",
      "\n",
      "Epoch 00036: ReduceLROnPlateau reducing learning rate to 1e-05.\n",
      "\n",
      "Epoch 00036: val_loss did not improve from 0.02978\n",
      "Epoch 00036: early stopping\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(train_images, \n",
    "                    train_labels,\n",
    "                    batch_size = 256,\n",
    "                    validation_split=0.15,\n",
    "                    epochs = 100,\n",
    "                    shuffle = True,\n",
    "                    callbacks= callbacks)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "80ed47b3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Loss\n",
    "plt.figure(figsize = (9, 6))\n",
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('U-Net model loss', fontsize = 15)\n",
    "plt.ylabel('loss', fontsize = 15)\n",
    "plt.xlabel('epoch', fontsize = 15)\n",
    "plt.legend(['train', 'validation'], loc='upper right')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "1619fbed",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Accuracy\n",
    "plt.figure(figsize = (9, 6))\n",
    "plt.plot(history.history['accuracy'])\n",
    "plt.plot(history.history['val_accuracy'])\n",
    "plt.title('U-Net model train accuracy', fontsize = 15)\n",
    "plt.xlabel('epoch', fontsize = 15)\n",
    "plt.ylabel('accuracy', fontsize = 15)\n",
    "plt.legend(['train accuracy', 'test accuracy'], loc = 'upper right')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "b7b11c4e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "313/313 [==============================] - 1s 3ms/step - loss: 0.0264 - accuracy: 0.9919\n",
      "\n",
      "Test accuracy: 0.9919000267982483\n"
     ]
    }
   ],
   "source": [
    "# Evaluating the model using the test set\n",
    "loss, accuracy = model.evaluate(test_images, test_labels, verbose = 1)\n",
    "print('\\nTest accuracy:', accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "4f1dc2a9",
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = model.predict(test_images)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "7484ef53",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_images = train_images.reshape(train_images.shape[0], w, h, 1)\n",
    "test_images = test_images.reshape(test_images.shape[0], w, h, 1)\n",
    "\n",
    "\n",
    "# Cifar's label need to reshape to scalar array\n",
    "test_labels = test_labels.reshape(10000, )\n",
    "\n",
    "\n",
    "\n",
    "# Helper functions to plot images \n",
    "def plot_image(i, predictions_array, true_label, img):\n",
    "  predictions_array, true_label, img = predictions_array, true_label[i], img[i]\n",
    "  plt.grid(False)\n",
    "  plt.xticks([])\n",
    "  plt.yticks([])\n",
    "\n",
    "  plt.imshow(img, cmap=plt.cm.binary)\n",
    "\n",
    "  predicted_label = np.argmax(predictions_array)\n",
    "  if predicted_label == true_label:\n",
    "    color = 'blue'\n",
    "  else:\n",
    "    color = 'red'\n",
    "\n",
    "  plt.xlabel(\"{} {:2.0f}% ({})\".format(classess[predicted_label],\n",
    "                                100*np.max(predictions_array),\n",
    "                                classess[true_label]),\n",
    "                                color=color)\n",
    "\n",
    "def plot_value_array(i, predictions_array, true_label):\n",
    "  predictions_array, true_label = predictions_array, true_label[i]\n",
    "  plt.grid(False)\n",
    "  plt.xticks(range(10))\n",
    "  plt.yticks([])\n",
    "  thisplot = plt.bar(range(10), predictions_array, color=\"#777777\")\n",
    "  plt.ylim([0, 1])\n",
    "  predicted_label = np.argmax(predictions_array)\n",
    "\n",
    "  thisplot[predicted_label].set_color('red')\n",
    "  thisplot[true_label].set_color('blue')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "88265776",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x1000 with 30 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the first X test images, their predicted labels, and the true labels.\n",
    "# Color correct predictions in blue and incorrect predictions in red.\n",
    "num_rows = 5\n",
    "num_cols = 3\n",
    "num_images = num_rows*num_cols\n",
    "plt.figure(figsize=(2*2*num_cols, 2*num_rows))\n",
    "for i in range(num_images):\n",
    "  plt.subplot(num_rows, 2*num_cols, 2*i+1)\n",
    "  plot_image(i, predictions[i], test_labels, test_images)\n",
    "  plt.subplot(num_rows, 2*num_cols, 2*i+2)\n",
    "  plot_value_array(i, predictions[i], test_labels)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "32d45153",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.16 (main, Jan 11 2023, 16:16:36) [MSC v.1916 64 bit (AMD64)]"
  },
  "vscode": {
   "interpreter": {
    "hash": "7789ff83f194acf566cc5f743b97f52059d095d5da8c0625de7b48fea8fb60ab"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}