Brizar

I tried a long time to predict the keypoints on a custom dataset, cleaned up the code, modified my depth images to be as close to the ITOP_side ones as...

Does torch have cuda access? e.g. run in environment: `python -c 'import torch;print(torch.cuda.is_available())'` What is your cuda-toolkit and torch version?

The above comment did not work for me, but the same procedure with the other github did work! The model now runs on every image, but the JSON Serializer cannot...

For me there are still two questions open: * Do we really want to use `np.float32` or ist `float` or `np.float64` better? There is one PR [1](https://github.com/MVIG-SJTU/AlphaPose/pull/1110) in this repo...

In the end, the ``install.bat`` wants to install the git project. But I struggled with the automatic version as well. It seems we should run everything from the parent directory....

I know that I had the same issues while using pip install torchreid without docker. But I solved it similarly. The following code in the requirements.txt is commented out: https://github.com/KaiyangZhou/deep-person-reid/blob/566a56a2cb255f59ba75aa817032621784df546a/requirements.txt#L18-L24...

Hello, I think you missunderstood the goal of this package. This package is for "person-reid" - person re-identification. Therefore using visual Neural-Nets to compress information about people and re-identifying those...

Should be the same fix as in https://github.com/leonid-pishchulin/poseval/pull/53 using ```pip install git+https://github.com/svenkreiss/poseval.git``` works for me.

I received a negative MOTA too, even though most other values are correct. Did you manage to fix this issue for your files?