search

ZiqiaoPeng / SyncTalk

[CVPR 2024] This is the official source for our paper "SyncTalk: The Devil is in the Synchronization for Talking Head Synthesis"
https://ziqiaopeng.github.io/synctalk/
Other
1.2k stars 138 forks source link

[WARN] Failed to load optimize #174

Open WinterJack002 opened 2 months ago

WinterJack002 commented 2 months ago

When I configure it according to the tutorial, and I run python main.py data/May --workspace model/trial_may -O --test --test_train --asr_model ave --portrait --aud ./demo/test.wav the code to inference with my own audio, but the output in terminal show :"[WARN] Failed to load optimizer.", why it happened??

Further more, the output result is terrible!(the entire facial area is lost, resulting in a large cavity) I think it is related to "Failed to load optimize". @ZiqiaoPeng ![Uploading bad result.jpg…]()

Sprinkle0 commented 1 month ago

I also encountered this problem, have you fixed it? @WinterJack002

WinterJack002 commented 1 month ago

Sorry for the late reply. I haven't solved it yet and have given up on this project now, but I'm keeping focusing on other similar works in face reconstruction and animation. Hope you can succeed and looking forward to your answer. Thank you.