GeneFace++ 通常依赖 CUDA 11.x
验证:
nvcc -Vconda create -n geneface python=3.8
conda activate geneface根据你 CUDA 版本选择命令,例如 CUDA 11.8:
pip install torch==1.13.1+cu117 torchvision==0.14.1+cu117 --extra-index-url https://download.pytorch.org/whl/cu117git clone https://github.com/yerfor/GeneFacePlusPlus.git
cd GeneFacePlusPluspip install -r requirements.txtpip install pywin32若报错:
pip install "numpy<1.24" scikit-image==0.19.3Windows 下 dlib 易失败,建议:
conda install -c conda-forge dlib
pip install face-alignmentpip install librosa soundfile moviepy如 moviepy 报错:
pip install imageio imageio-ffmpegGeneFace++ 需要:
参考官方 README:
bash scripts/download_pretrained.shWindows 无 bash,可手动下载并放到:
checkpoints/python inference.py --pgvs_mode fast --a2m_model_path checkpoints/audio2motion --m2v_model_path checkpoints/motion2video --inp audio/test.wav --out output.mp4如果你在 Windows 上遇到太多坑,强烈建议:
如果你愿意,我可以:
你现在是准备训练还是只推理?