search

yerfor / GeneFacePlusPlus

GeneFace++: Generalized and Stable Real-Time 3D Talking Face Generation; Official Code
MIT License
1.57k stars 225 forks source link

How to improve the output video quality? #199

Closed gokula-krishna-dev closed 3 months ago

gokula-krishna-dev commented 6 months ago

I just set up the project and tried to run inference on May's video.

I trained the video using https://github.com/yerfor/GeneFacePlusPlus/blob/main/egs/datasets/May/lm3d_radnerf_sr.yaml & https://github.com/yerfor/GeneFacePlusPlus/blob/main/egs/datasets/May/lm3d_radnerf_torso_sr.yaml.

I ran the following commands

CUDA_VISIBLE_DEVICES=0 python tasks/run.py --config=egs/datasets/May/lm3d_radnerf_sr.yaml --exp_name=motion2video_nerf/May_head --reset
CUDA_VISIBLE_DEVICES=0 python tasks/run.py --config=egs/datasets/May/lm3d_radnerf_torso_sr.yaml --exp_name=motion2video_nerf/May_torso --hparams=head_model_dir=checkpoints/motion2video_nerf/May_head/model_ckpt_best.pt --reset

For inference, I ran

CUDA_VISIBLE_DEVICES=0  python inference/genefacepp_infer.py --head_ckpt= --torso_ckpt=checkpoints/motion2video_nerf/may_torso --drv_aud=data/raw/val_wavs/MacronSpeech.wav

https://github.com/yerfor/GeneFacePlusPlus/assets/33853819/6fd984f8-4613-4474-adeb-d04721dbbe38