https://github.com/masc-it/yolov5-api-rust

Rust API to run predictions with YoloV5 models.
https://github.com/masc-it/yolov5-api-rust

actix-web onnx rust yolov5

Last synced: 12 days ago
JSON representation

Rust API to run predictions with YoloV5 models.

• Host: GitHub
• URL: https://github.com/masc-it/yolov5-api-rust
• Owner: masc-it
• License: mit
• Created: 2022-07-16T15:20:21.000Z (almost 2 years ago)
• Default Branch: main
• Last Pushed: 2022-08-04T20:51:09.000Z (almost 2 years ago)
• Last Synced: 2024-02-29T05:33:40.070Z (4 months ago)
• Topics: actix-web, onnx, rust, yolov5
• Language: Rust
• Homepage:
• Size: 26.4 KB
• Stars: 26
• Watchers: 1
• Forks: 4
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Lists

• awesome-yolo-object-detection - masc-it/yolov5-api-rust - it/yolov5-api-rust?style=social"/> : Rust API to run predictions with YoloV5 models. (Other Versions of YOLO)

README

        
# YoloV5-API [WIP]

API to run inferences with YoloV5 models. Written in Rust, based on OpenCV 4.5.5

If you need a C++ version, check my [C++ Yolov5-API](https://github.com/masc-it/yolov5-api-cpp)

## Requirements

- [OpenCV 4.5.5](https://github.com/opencv/opencv/releases/tag/4.5.5) installed on your system

  - Follow [Rust opencv README](https://github.com/twistedfall/opencv-rust)

## Model config

**Data** directory must contain your config.json

**config.json** defines:

- ONNX absolute model path

- input size (640 default)

- array of class names

A dummy example is available in the _data/_ folder

## Docker

    docker run --name rust-yolov5-api -v :/app/data -p 5000:5000 mascit/rust-yolov5-api:latest

## Build

Development:

    cargo build

Release:

    cargo build --release

## Run

    cargo run

For Windows users: Assure to have _opencv_world455.dll_ in your exe directory.

# Endpoints

## /predict [POST]

### Body

- Image bytes (binary in Postman)

### Headers

- X-Confidence-Thresh

  - default 0.5

- X-NMS-Thresh

  - default 0.45

- X-Return

  - image_with_boxes

    - A JPEG image with drawn predictions

  - json (default)

    - A json array containing predictions. Each object defines: xmin, ymin, xmax, ymax, conf, class_name