Hei Law
Yeah, that sounds better. I think we can add an option to here: https://github.com/DLR-RM/BlenderProc/blob/ac97e4ac380764d47dbb0a214a1758d1cc6cb1d2/blenderproc/python/renderer/RendererUtility.py#L470. Then we can find a way to redirect all the output generated by this function: https://github.com/DLR-RM/BlenderProc/blob/ac97e4ac380764d47dbb0a214a1758d1cc6cb1d2/blenderproc/python/renderer/RendererUtility.py#L511
Sounds good. I am going to try it out. If it works, I will submit a pull request.
https://github.com/princeton-vl/CornerNet-Lite/issues/76#issuecomment-495322842 When you evaluate on the test set, you will need to submit the results to the COCO evaluation server as the ground-truths of the test set are not available....
I only tested it with CUDA10 so I am not sure if it would run on CUDA9. I would recommend upgrading your driver if possible.
Can you include the complete error message?
On our machine with a 1080Ti and Core-i7 7700k, the average FPS of CornerNet-Squeeze is 33fps while that of YOLOv3 is 30fps. The difference in the inference time can be...
The speed can also depend on the frequency of CPU since part of the code is executed on CPU. However, I am not sure how much it would affect the...
COCO does not provide annotation for the test set. That's why the COCO evaluation API gives -1, when you are evaluating the detector on the test set. You need to...
You will need to also download the [pretrained model](https://drive.google.com/file/d/1MQDyPRI0HgDHxHToudHqQ-2m8TVBciaa/view?usp=sharing) for CornerNet-Saccade and put it under `cache/nnet/CornerNet_Saccade/`