JohnPPinto/food-geek-food-image-segmentation

This repo contains all the code for food image segmentation - Food Geek App.

Food Geek: A Food Segmentation Application

A Food Segmentation project, built using Food 101 Dataset, YOLOv8, ONNX, FastAPI, Docker and Gradio.

You can find the live and running web application over here and the project documentation over here.

Local Installation

1. Docker Method:

• If you prefer to use a Docker Image, you can pull the image from the docker hub by using the below command:

docker pull johnppinto/food-geek:0.2.0

• Once you have the docker image in your system you can use this command to run the web application on the local host port 7860 (http://localhost:7860/).

docker run -it --rm -p 7860:7860 johnppinto/food-geek:0.2.0

2. Local Build Method:

• Clone this repository and set the current working directory to demo directory.

git clone https://github.com/JohnPPinto/food-geek-food-image-segmentation.git
cd food-geek-food-image-segmentation/demo

• Execute the requirements file in your environment.

pip install -r requirements.txt

• Now you can run the app.py file.

python app.py

The web application will run on the following address: http://127.0.0.1:8000, by default uvicorn server uses this address to run any application.