search

fudan-generative-vision / hallo

Hallo: Hierarchical Audio-Driven Visual Synthesis for Portrait Image Animation
https://fudan-generative-vision.github.io/hallo/
MIT License
9.49k stars 1.3k forks source link

sync-c和sync-d评价指标如何计算 #182

Closed hnsywangxin closed 3 months ago

hnsywangxin commented 3 months ago

感谢作者的开源！ 目前我看了论文和代码，好像没有讲这2个指标是如何计算的，论文也没看到引用的出处，在网上也没看到相关指标计算的方式。 我个人猜测，咱们的这两个计算方式应该是跟wav2lip论文中的lse-c和lse-d指标是一致的。如果我理解的有问题，恳请作者指导一下如何计算这两个指标。 谢谢！

crystallee-ai commented 3 months ago

Thanks for your attention! We refer to the implementation of this repo: https://github.com/joonson/syncnet_python

hnsywangxin commented 3 months ago

Thanks for your attention! We refer to the implementation of this repo: https://github.com/joonson/syncnet_python

了解了，谢谢