i-l-sokolov / Flask_deployment_of_object_detection_pytorch_NN

Deployment of a pretrained neural network for object detection using Flask

Object Detection with Pretrained Neural Network

This project is a deployment of a pretrained neural network for object detection on images. It was used pretrained Resnet50_FPN

Requirements

All the necessary dependencies are listed in the environment.yml file. You can install them via conda:

conda env create -f environment.yml

Demo

NN detection of objects
Developed web interface for NN usage

GIF with demo could be found by link

Usage

Activate the conda environment:

conda activate <env_name>

Navigate to the folder with the app.py file. Run the Flask app:

python app.py --port <port_number>

By default, the app runs on port 8000. You can specify a different port by setting the --port flag.

Navigate to 127.0.0.1:<port_number> in your web browser to use the web interface for object detection.