Closed xinzi2018 closed 4 years ago
Our style transfer framework deals with a fixed skeleton. To test our pretrained model, you need to first retarget your BVH to the skeleton described in style_transfer/global_info/rest.bvh. We'll provide more information on how to train the model with your own customized skeleton in the near future. Hope this would help.
in style transfer ,when i want to change content BVH, how can i do? you say the skeleton should same as style_transfer/global_info/rest.bvh, how can we get it from our own avis?
Extracting 3D motions from 2D videos (3D motion reconstruction/pose estimation) is a fundamental research problem in computer vision. And it's beyond the scope of our style transfer framework. We only support direct usage of 2D videos as the style input.
thanks a lot !
In styleTransfer,when i use other BVHs,why cant run the style_transfer/demo.sh successfully? is my skeleton in my BVH is lack? the skeleton must match it that you give?