animegan2-pytorch

animegan2-pytorch copied to clipboard

mypfp.io, an anime-like NFT avatar generator service

10

Thanks for publishing a very useful model and sample code. We've been using your "Face Portrait v2 model" in our experimental NFT project [mypfp.io](https://mypfp.io), an anime-like NFT avatar generator service....

dressedanonymous

No matter what image I try I get this following error: MessageError: RangeError: Maximum call stack size exceeded. I’m running on google colab and no matter what I do I...

Joeyd348039

zlwnge

文章介绍了模型怎么使用，https://mp.weixin.qq.com/s/9FRoRZNJQEFwfPSxu1kSPg 但从效果看，图片结果并不是很理想

DataShare-duo

Changing model params?

2

Is there any way to change the parameters (kernel width, step size, padding), of the NN blocks defined in model.py?

bhughesshelton

Training details

1

Hi! Great work! May I ask what GAN loss are you using? I'm trying to achieve a similar effect with WGAN, but the result is quite poor. Also, you mentioned...

chrisha97

3d face reconstruction

1

Any suggestion for 2d to 3d image reconstruction? I was lookin at [This](https://github.com/microsoft/Deep3DFaceReconstruction) project by MS seems to be trained on actual faces, maybe it could work applied to the...

loretoparisi

I add a video processor to it. hope you can merge it. Thank you.

oakboat

之前做过可视化，用的2次样条插值法，能够平滑些，可以试试

DataShare-duo

ResizeRight and Antialiased CNNs recommendation

1

Rather than F.interpolate with optional alignment of corners you could try [ResizeRight](https://github.com/assafshocher/ResizeRight) and in places where convolution strides are above 1 you can add [Antialiased CNNs](https://github.com/adobe/antialiased-cnns)'s BlurPool after them and...

torridgristle