YOLOv7 Live Detection Application on Node.js. Serving YOLOv7 in Node.js using onnxruntime-node.

git clone https://github.com/Hyuto/yolov7-node.git
cd yolov7-node
yarn install # Install dependencies

yarn start

YOLOv7 model converted to onnx model.

used model : yolov7-tiny
size       : 24 MB

⚠️ Expensive Computation : YOLOv7 model used in this repo is the smallest with size of 24 MB, so other models is definitely bigger than this which can exhausting computation.

Use another YOLOv7 model.

1. Clone yolov7 repository

   git clone https://github.com/WongKinYiu/yolov7.git && cd yolov7

   Install requirements.txt first

   pip install -r requirements.txt

   Then export desired YOLOv7 model and configurations to onnx

   python export.py --weights <YOLOv7-MODEL>.pt --grid --end2end --simplify \
        --topk-all 100 --iou-thres 0.65 --conf-thres 0.35 --img-size 640 640 --max-wh 640

   Note : You can run it on colab too

   

2. Copy yolov7*.onnx to ./src

3. Update modelInfo in index.js to new model name

   ...
   // model configs
   const modelInfo = {
      name: "yolov7-tiny.onnx", // change this to new model filename
      inputShape: [1, 3, 640, 640], // change this if model input shape isn't 640 x 640
   };
   ...

4. Done! 😊

https://github.com/WongKinYiu/yolov7