takeshi iwa
takeshi iwa
I would appreciate it if you could add explanation to use other devices such as raspi camera and USB camera.
Hi, just sharing. I created GitHub Actions script to build libedgetpu for Windows. https://github.com/iwatake2222/InferenceHelper_Binary/blob/master/.github/workflows/build_edgetpu_windows.yml
Thank you for trying my project, but I did it more than 5 years ago. So, I don't remember it and can't really help you. Sorry. The only thing that...
hmm, I can't think of any other cause.
I'm not sure because I don't have Jetson Orin, but it may work.
This issue was closed because it has been inactive.
> about int64 You can ignore this. > about issues when converted to FP16 Comment the following line out. https://github.com/iwatake2222/InferenceHelper/blob/7c975cf9d32016aa6bfceabf6922fe6d808dc5eb/inference_helper/inference_helper_tensorrt.cpp#L44 > output The window doesn't output anything causes this.
> Why is this the case,I noticed that FGR is nine images displayed together,but PHA is displayed sequentially, why? I'm not familiar with details of model outputs. You can ask...
First, I recommend trying `pj_tensorrt_cls_mobilenet_v2` with `USE_INT8_WITH_CALIBRATION` enabled. So that you can see working INT8 calibration. Then, you can find out differences between `pj_tensorrt_cls_mobilenet_v2` and your code/settings.
This issue was closed because it has been inactive.