nagadomi / waifu2x

Image Super-Resolution for Anime-Style Art
http://waifu2x.udp.jp/
MIT License
27.56k stars 2.71k forks source link

how to deal with variable input #228

Open ghost opened 6 years ago

ghost commented 6 years ago

if the image 's size are variable,the output image's size are variable too,so how to build this nn

nagadomi commented 6 years ago

Fully Convolutional Networks can handle variable input size. And, waifu2x uses the fixed block conversion algorithm like Tiled rendering in 3DCG field, it is particularly aimed at reducing GPU memory usage. tiled render