新开源 免训练 一张图就可以换视频

xiaoxin

星级打分

• 1
• 2
• 3
• 4
• 5

平均分:NAN  参与人数:0  我的评分:未评

作者的开源模型是128*128的，所以脸部比较模糊，刚开源所以效果会越来越好的。

https://github.com/s0md3v/roop

fanda

测试了，可以运行，效果不太满意

fanda

4G显存笔记本可以用，不过cpu和gpu都爆满

rinlinton

fanda 发表于 2023-5-30 17:39
4G显存笔记本可以用，不过cpu和gpu都爆满

效果怎麼樣

xiaoxin

rinlinton 发表于 2023-5-30 17:52
效果怎麼樣

脸部有点糊，我觉得效果可以，作者没有开源更高的模型。

seancai110

xiaoxin 发表于 2023-5-30 18:31
脸部有点糊，我觉得效果可以，作者没有开源更高的模型。

感觉和小程序里的照片换脸，还是差距有点大啊。小程序里的，应该是给每个视频做过专门优化吧。

perryfans

记录了

lijie207

perryfans 发表于 2023-5-30 19:51
记录了

      error: Microsoft Visual C++ 14.0 or greater is required. Get it with "Microsoft C++ Build Tools": https://visualstudio.microsoft.com/visual-cpp-build-tools/
      [end of output]

  note: This error originates from a subprocess, and is likely not a problem with pip.
  ERROR: Failed building wheel for insightface
Failed to build insightface
ERROR: Could not build wheels for insightface, which is required to install pyproject.toml-based projects

这个错误怎么处理啊 ？

lijie207

      error: Microsoft Visual C++ 14.0 or greater is required. Get it with "Microsoft C++ Build Tools": https://visualstudio.microsoft.com/visual-cpp-build-tools/
      [end of output]

  note: This error originates from a subprocess, and is likely not a problem with pip.
  ERROR: Failed building wheel for insightface
Failed to build insightface
ERROR: Could not build wheels for insightface, which is required to install pyproject.toml-based projects
出现这个错误，该怎么处理啊？

xiaoxin

lijie207 发表于 2023-5-30 19:59
error: Microsoft Visual C++ 14.0 or greater is required. Get it with "Microsoft C++ Build Tool ...

Installed ffmpeg and added to path(google it) and then installed K-lite codec pack.

Installed my graphics card specific latest CUDA toolkit and CuDNN from NVIDIA website.

Installed Build Tools for Visual Studio 2017 (version 15.9) from https://my.visualstudio.com/Down ... com~older-downloads

Now clone the repo (git clone https://github.com/s0md3v/roop)

cd roop && pip install -r requirements.txt

Downloaded https://drive.google.com/file/d/ ... view?usp=drive_link and kept it in roop directory.

pip uninstall onnxruntime onnxruntime-gpu

Now install torch with cuda (pip install torch==2.0.0+cu118 torchvision==0.15.1+cu118 torchaudio==2.0.1 --index-url https://download.pytorch.org/whl/cu118) It will uninstall all previously installed torch stuff before installing it

pip install onnxruntime-gpu

That's all. Now you can run with gpu flag (python run.py --gpu)