Welcome to new followers who arrived from here!
— Alex J. Champandard 🌱 (@alexjc) 26 février 2025
FYI: the code in 2016 used about 80Mb of statistics (convolution kernels) learned by training on ImageNet dataset. I since figured out how to make it work with random kernels. The style images used were out-of-Copyright. https://t.co/lZPj08zpES
Welcome to new followers who arrived from here! FYI: the code in 2016 used about 80Mb of statistics (convolution kernels) learned by training on ImageNet dataset. I since figured out how to make it work with random kernels. The style images used were out-of-Copyright.