About 3*3*1*1 conv layer in encoder

naoto0804 / pytorch-AdaIN

Unofficial pytorch implementation of 'Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization' [Huang+, ICCV2017]

MIT License

1.08k stars 208 forks source link

Open diaodeyi opened 3 years ago

diaodeyi commented 3 years ago

Thanks for your code, I want to know what do you think of adding this layer as the original author did. Is the experimental result?