First time here? Check out the FAQ!
onnx model shows slower inference when setting cuda as backend I have a net loaded from onnx: net = cv2.dnn.readNetFrom