YoungSeng / QPGesture

QPGesture: Quantization-Based and Phase-Guided Motion Matching for Natural Speech-Driven Gesture Generation (CVPR 2023 Highlight)
82 stars 7 forks source link

Same result for different video #23

Open ylhua opened 4 months ago

ylhua commented 4 months ago

Thanks for your great work and code! I've tried your method and tested on different audio without text, following the tutorial. However, I found all the animated results were same. Here are the results from two audio files.

https://github.com/YoungSeng/QPGesture/assets/37425013/e81b391d-4f54-414b-824c-df27403d74bc