Skip to content

Latest commit

 

History

History
54 lines (45 loc) · 1.85 KB

README.md

File metadata and controls

54 lines (45 loc) · 1.85 KB

ImagePredService

Django Web Service with Tensorflow Image prediction

Installation

Framework

Python 3.6
$ pip install django
$ pip install djangorestframework

Machine Learning

$ pip install tensorflow==2.0.0
$ pip install keras==2.1.5
$ pip install pillow==7.0.0

Make sure the Tensorflow version 2.0.0 is installed. Other versions cause different errors.
New TF versions could be more stabe in the future (Today: 29.01.2020).

Usability

Model Training

Check the tutorials to train your model.
Create two files "labels.txt" and "saved_model.pb" and copy into this project.

Run Django

Install the prerequisites.
Run in virtual environment.
$ cd "cloned directory"
$ python manage.py runserver
Settings for postman:
Postman
Recreate Postman post request in your application.
Response:
"predicted" - label from "labels" with highest probability
"defect" - first entry in "labels" and its probability
"ok" - seccond entry in "labels" and its probability
"robotarm" - third entry in "labels" and its probability
"id" - ID fron Django DB
"file" - Link to file

Additional informations

There is no picture segmentation only picture classification.
The uploaded picture will get an ID.
Only the current picture exists in Database and File Server.
Old picture will be deleted when uploading the new picture to save space on server.

Links to Tutorials

Rest Framework

https://www.techiediaries.com/django-rest-image-file-upload-tutorial/

Machine Learning

Tensorflow Model was trained using this sources. The Model was trained in Colab (Google Drive).
https://colab.research.google.com/github/tensorflow/hub/blob/master/examples/colab/tf2_image_retraining.ipynb#scrollTo=FlsEcKVeuCnf https://github.com/tensorflow/hub

Deployment to PYTHONANYWHERE

https://www.youtube.com/watch?v=Y4c4ickks2A