search

chenzhike110 / Semantic-Motion-Retargeting

Code of [CVPR 2024] "Semantics-aware Motion Retargeting with Vision-Language Models"
MIT License
28 stars 2 forks source link

dataset and training #4

Open guosongrui opened 2 months ago

guosongrui commented 2 months ago

Hello author, thank you for your excellent contribution. According to the operation instructions in your README, I have read the code files fbx2bvh.py, fbx2data.py, and train.py. I noticed that in train.py, the data such as weight, vgrp_labels, coords, faces, uv_coords, and joints_origin that you mentioned are not directly used. At the same time, I am also puzzled about what the specific data format should look like. Could you provide a more complete example of the data? This would be helpful for training the model.

chenzhike110 commented 2 months ago

We will add preprocessing and data-related documentation this week.

guosongrui commented 2 months ago

We will add preprocessing and data-related documentation this week.

Thank you very much for your reply