Vegetebird / MHFormer

[CVPR 2022] MHFormer: Multi-Hypothesis Transformer for 3D Human Pose Estimation
MIT License
530 stars 85 forks source link

How to visualize multi-head attention maps? #85

Closed TomCatOnline closed 1 year ago

TomCatOnline commented 2 years ago

Thanks for your great work! I wonder how to visualize multi-head attention maps as you do in fig12&13 ? Looking forward to your reply!

Vegetebird commented 1 year ago

You can use the code in https://github.com/Vegetebird/StridedTransformer-Pose3D/blob/main/common/plot_att.py