Custom Image Classfier - Flask Deployment

Model: Resnet50 + LinearSVC

• example_model folder contains classifier model for model testing

• uploads folder contain images for testing the classifier model

• classifier_model folder contained trained classifier model based on user input classes

• images folder contains .p file for model training based on user input classes

Usage

Run app.py using below command to start Flask API

python app.py

1. model builder

Pipeline

1. Scraping google images based on user input classes
2. Extract feature by using resnet50 and store data in images folder
3. Trained model with LinearSVC
4. Return an image classifier model file that can be opened with joblib

2. predict image via uploading

Upload image and submit

3. predict image via image url

Predict image from image link address

4. build image classifier model trained on collected data

run train_model_from_collected_data.py