Results 55 comments of ZJYang

Thank you for your valuable suggestions! We believe that using face landmarks may not be the optimal method for controlling face motion, as it could introduce additional noise.

I'm not sure the specific situation. May be the max value of sampe_idx is len(dataset_len) - 1?

Seems the hugging face service response timed out according to the log.

Thank you for your interest in our work! We appreciate your viewpoints. 1. We utilize optical flow to create intermediary frames for reducing AniPortrait's time consumption, and the artifacts are...

这部分代码还在优化,后续有进展会在该rep中同步。谢谢关注!

@DeepAichemist 是的,训练模型时采用的尺寸是512*512,推理时缩小视频尺寸会导致表情精度下降。

Perhaps the pose video is too long. Consider using the command -L 300 to obtain a shorter video.

https://huggingface.co/spaces/ZJYang/AniPortrait_official We recommend you to use our gradio demo on hugging face.

> the demo on hugging face doesn't seem to work? I've been trying for two days but it just gives an error after a few hours We have updated a...

可以尝试下载所有依赖预训练模型的完整文件夹。 https://github.com/Zejun-Yang/AniPortrait/blob/195652d07be03762d47d5a43d8a9ca5e7a0fe233/README.md?plain=1#L94