Peter

The file of .mexw64 only run in windows system. But your system is win10. I propose that You recompile Mex_ZBufferC.cpp in your os environment。You can get the file of Mex_ZBufferC.cpp...

Have you solved this problem? Why is my output image reversed? The generated picture is as follows  the orignial pose image  and I use 68 key points to...

it work well when I use "copyfile( 'Out/pncc_code.mat','3ddfa/pncc_code.mat')" , my matlab version is R2021b

> https://github.com/eriklindernoren/PyTorch-GAN/blob/a163b82beff3d01688d8315a3fd39080400e7c01/implementations/bicyclegan/models.py#L153 > > in_channels -> 频道 （我猜） thank you for your solution! it works well

> @AtlantixJJ你能试试最近的 pytorch（我用 >=1.3 测试过）和 CUDA 10 吗？ Excuse me, my GPU is NVIDIA rtx3090. I only use cuda11. How can I solve this problem.

How can I get the voxceleb1 dataset. the method in project of https://github.com/AliaksandrSiarohin/video-preprocessing has become invalid.

> 我可以尝试在 OneDrive 上上传 voxceleb1，但我不确定这是否可行，因为数据集太大。如果它有效，我将更新自述文件。 thank you for your reply。

Thank you for your work. There is no doubt that you have done your best.I have replaced the dataset and it can work.

> 这是输出和错误日志 have you solved this problem?

you can not use pip to install pytorch3d,you should use conda as follows conda install pytorch3d=0.6.0 -c pytorch3d