Wenjia Wang
Wenjia Wang
In [DOC](https://github.com/open-mmlab/mmhuman3d/blob/main/docs/preprocess_dataset.md#vibe), it is said that the preprocessed `mpi_inf_3dhp` and `pw3d` npz files from SPIN. The files will be used in [VibeConverter](https://github.com/open-mmlab/mmhuman3d/blob/main/mmhuman3d/data/data_converters/vibe.py#L33-L36). However, when I have run [fetch_data.sh in spin](https://github.com/nkolot/SPIN/blob/master/fetch_data.sh),...
- Add new model classes when ready. [TODO](https://github.com/open-mmlab/mmhuman3d/blob/main/mmhuman3d/core/visualization/renderer/torch3d_renderer/meshes.py:L43) - rewrite the TexturesUV `join_scene` or add a compatible function for textures slice. Add rewrite the functions [__getitem__](https://github.com/open-mmlab/mmhuman3d/blob/main/mmhuman3d/core/visualization/renderer/torch3d_renderer/meshes.py:L381) and [join_meshes_as_scene](https://github.com/open-mmlab/mmhuman3d/blob/main/mmhuman3d/core/visualization/renderer/torch3d_renderer/meshes.py:L482).
- Need to add a smplx_uv.npz like smpl_uv.npz. Solved the [TODO](https://github.com/open-mmlab/mmhuman3d/blob/main/mmhuman3d/core/visualization/renderer/torch3d_renderer/uv_renderer.py:L45). It needs to be put in data/body_models/smplx - Need to solved the sample bug when the resolution is too...
I want to train on my video dataset
hello, seems that pytorch3d cannot save mesh with TexturesVertex. Is that so? [save_obj](https://github.com/facebookresearch/pytorch3d/blob/main/pytorch3d/io/obj_io.py#L661-L670)
Hi there @AliaksandrSiarohin , thanks for your wonderful work. Could you share your pre-trained models? I only find the moving-gif model.
Dear author, have you visualized the attention masks like the pics in the original paper? Can that really find the location of each character? 
I have not found **MeanBlur()** in common.py
Please write your suggestions in below comments. I will rearrange them and summarize a plan list. ## Cameras. - [ ] Add verbose switch in renderer. ## Renderer. - [...
This PR will last for weeks.