Xianghui

yes, we are preparing the release of our training and test code, it will come in the following weeks. stay tuned :-)

Hello, Thanks for your interest. The data and registration at 30fps are still under processing. There are significantly more data when annotating at 30fps, so we need a bit more...

Code for BEHAVE model can be accesed here: https://github.com/bharat-b7/BEHAVE

hi, the full video sequence was released in [our website](https://virtualhumans.mpi-inf.mpg.de/behave/license.html), and you can find the example code to extract frames from video [here](https://github.com/xiexh20/behave-dataset#generate-images-from-raw-videos).

Hi, I did not try RTX30xx devices but I think the env should work with cuda 11.0, but with a lower torch version. Maybe you can try to reinstall the...

Hi, Thank you for your interest. Do you mean detailed finger pose? Unfortunately, our capture glove was broken and the hand pose is not usable :( Current hand pose optimized...

> > Hi, I think the problem may be due to the submodule `LoFTR` is not properly installed. You can install it by running `git submodule update --init --recursive`. >...

I also have the same problem. Any updates? thanks!

Problem solved, they can be downloaded with wget directly.