semchan
Glitchy output
My outputs are all shakey and the lips do not track properly.
Even the provided example saved in the git is glitchy, what's going on?
https://github.com/semchan/HyperLips/raw/main/results/result_voice.mp4
For anyone wanting to try for themselves but don't want the hassle of setting it all up, I made a colab which can be found in my fork:
https://github.com/anothermartz/Easy-HyperLips
You can find another example of the glitchy lips saved within the colab.
Also having issues, even for video/audio pairs from the same speaker in mead! it works perfectly for the same audio/video sample though. Is the Inference_hyperlips.py the correct thing to be running for a lipsync use case?
https://github.com/semchan/HyperLips/assets/4791204/77c70de7-8166-4f9f-94d5-b0de0de1a1d7
Same problem here. Did anybody find a solution for that? Would be nice to at least replicate the results shown in the demo video.
