This repository is a small website together with a small backend that allows a user to upload an image and remove its background.
- Docker and docker-compose.
- The
Yolov8m-seg.onnxmodel stored in the/model/folder.
The goal of this repository is to build a small application using Rust. One of the important aspects for me was to use a Neural Network. The backend therefore consists of a Rust API that is linked to an ONNX Runtime which makes an inference with the Yolov8 Segmentation Model. The API framework I used is Axum. The frontend is built in Svelte.
There's an automated deployment action that is defined. This action simply removes any previous docker compose session and builds another up again. An important thing to note is to have the Yolo model stored somewhere and mount it correctly in the docker-compose.yml. Once that is done, you should be good to go.