Question about the evaluation for audio-to-dance?

Garfield-kh / TM2D

[ICCV 2023] TM2D: Bimodality Driven 3D Dance Generation via Music-Text Integration

75 stars 3 forks source link

Closed blue-blue272 closed 4 months ago

blue-blue272 commented 9 months ago

In Table1, how to compute the FID and Div? I think the processing can not produce keypoints3D for geometric features and kinetic features.

Garfield-kh commented 9 months ago

Hi, thank you for the interest! We following the evaluation code of Bailando.