Hallo2: Long-Duration and High-Resolution Audio-driven Portrait Image Animation,
2024/10/22
- 修复task地址绝对引用可能出现的问题。(fix bug)
- Currently, only square 512 images and 2x magnification are supported(目前仅支持方形512图像和2倍放大,官方模型和方法所限)
- input audio must be *.wav (输入的音频格式只能是wav,采样用的是16000,你要用高保真的自己合成就是了,别矫情.)
In the ./ComfyUI /custom_node directory, run the following:
git clone https://github.com/smthemex/ComfyUI_Hallo2
pip install -r requirements.txt
if using embeded comfyUI,in your "X:\ComfyUI_windows\python_embeded "(便携包的comfyUI用户在python_embeded目录下用以下命令安装)
python -m pip install -r requirements.txt
Possible installation difficulties that may be encountered(可能会遇到的安装难题):
- 2.1 audio-separator
- 2.1.1 If' pip install audio-separator' building wheel fail(diffq),makesure has install visual-cpp-build-tools in window
安装audio-separator可能会出现vs的报错,确认你安装了visual-cpp-build-tools - 2.1.2 Although there are ‘visual-cpp-build-tools’, it still fails(diffq). If you are using the ComfyUI portable package or Akiba package, please add the interpreter address to the Windows system variable.
虽然有‘visual-cpp-build-tools,但是还是失败(diffq),如果使用的是comfyUI便携包,或者秋叶包,请将解释器地址加入windows的系统变量里,Linux用户,你都用Linux了,就不用我教了吧,window的做法是,将X:\ComfyUI_windows\python_embeded 和F:\ComfyUI_windows\python_embeded\Scripts 2个地址加入Path系统变量里。 - 3.2 ffmpeg
- 3.3 If the module is missing, Remove the requirements' # symbol,please pip install
- 3.4 onnx 错误 少了啥,就去掉#号,重新安装
所有模型下载地址(all checkpoints):huggingface
├── ComfyUI/models/Hallo/
|-- audio_separator/
| |-- download_checks.json
| |-- mdx_model_data.json
| |-- vr_model_data.json
| `-- Kim_Vocal_2.onnx
|-- face_analysis/
| `-- models/
| |-- face_landmarker_v2_with_blendshapes.task # face landmarker model from mediapipe
| |-- 1k3d68.onnx
| |-- 2d106det.onnx
| |-- genderage.onnx
| |-- glintr100.onnx
| `-- scrfd_10g_bnkps.onnx
|-- facelib
| |-- detection_mobilenet0.25_Final.pth
| |-- detection_Resnet50_Final.pth
| |-- parsing_parsenet.pth
| |-- yolov5l-face.pth
| `-- yolov5n-face.pth
|-- hallo2
| |-- net_g.pth
| `-- net.pth
|-- motion_module/
| `-- mm_sd_v15_v2.ckpt
`-- wav2vec/
`-- wav2vec2-base-960h/
|-- config.json
|-- feature_extractor_config.json
|-- model.safetensors
|-- preprocessor_config.json
|-- special_tokens_map.json
|-- tokenizer_config.json
`-- vocab.json
Normal checkpoints
├── ComfyUI/models/
|-- upscale_models/
| `-- RealESRGAN_x2plus.pth
|-- vae/
| `-- vae-ft-mse-840000-ema-pruned.safetensors
|-- checkpoints/
| `-- v1-5-pruned-emaonly.safetensors # any sd1.5 if load fail, changge another
hallo2
@misc{cui2024hallo2,
title={Hallo2: Long-Duration and High-Resolution Audio-driven Portrait Image Animation},
author={Jiahao Cui and Hui Li and Yao Yao and Hao Zhu and Hanlin Shang and Kaihui Cheng and Hang Zhou and Siyu Zhu and️ Jingdong Wang},
year={2024},
eprint={2410.07718},
archivePrefix={arXiv},
primaryClass={cs.CV}
}