search

rlgnswk / NeRFFaceSpeech_Code

One-shot Audio-driven 3D Talking Head Synthesis via Generative Prior, CVPRW 2024
https://rlgnswk.github.io/NeRFFaceSpeech_ProjectPage/
MIT License
53 stars 2 forks source link

Please fix these issues in code / readme #4

Closed nitinmukesh closed 1 month ago

nitinmukesh commented 1 month ago

Some errors can be fixed, without that it will not work for Linux or Windows users

1. Fix path in code (it is currently hardcoded) [model_path = '/home/gihoon/NeRFFaceSpeech_CVPR/pretrained_networks/BFM_for_3DMM-Fitting-Pytorch/BFM/BFM09_model_info.mat']

NeRFFaceSpeech_Code\3DMM-Fitting-Pytorch\core__init__.py model_path = './pretrained_networks/BFM_for_3DMM-Fitting-Pytorch/BFM/BFM09_model_info.mat'

  1. Missing checkpoint (plz include in Readme) Download SadTalker_V0.0.2_256.safetensors https://github.com/OpenTalker/SadTalker/releases to NeRFFaceSpeech_Code\pretrained_networks\sad_talker_pretrained

  2. Missing file (plz include in Readme) Download https://huggingface.co/wsj1995/sadTalker/blob/af80749f8c9af3702fbd0272df14ff086986a1de/BFM09_model_info.mat to NeRFFaceSpeech_Code\pretrained_networks\BFM_for_3DMM-Fitting-Pytorch\BFM

  3. Please fix the inference command, incorrect path of audio file

python StyleNeRF/main_NeRFFaceSpeech_audio_driven_from_image.py \ --outdir=out_test --trunc=0.7 \ --network=pretrained_networks/ffhq_1024.pkl \ --test_data="test_data/test_audio/AdamSchiff_0.wav" \ --test_img="test_data/test_img/32.png"

rlgnswk commented 1 month ago

We've done. Thank you for sharing your experiences!

nitinmukesh commented 1 month ago

Thank you. I was about to raise PR but you are very fast.

rlgnswk commented 1 month ago

I have cited you at readme. check it!

nitinmukesh commented 1 month ago

Thank you @rlgnswk .