Convert .pt/.pth into onnx to use on opencv

[thoufeekx]

From the notebook 4 I have a .pt file and I need to use it in open cv so it can detect the objects (sushi, steak, pizza) on live and draw bounding box with labels.

When i did research on how to do it I realized that the file .pt need to convert into .onnx to use in opencv. The tutorial speaks about yolo model. So I dont know the specifics. I would love to use my trained model on opencv and detect live using videocapture().