Resources for image refinment

[Clorr]

A bunch of resources i'm trying to get into to explore image refinment:

An article by a guy from Let's Enhance about using VGG16: https://medium.com/lets-enhance-stories/content-and-style-loss-using-vgg-network-e810a7afe5fc

Another approach with feature loss: Generating Images with Perceptual Similarity Metrics based on Deep Networks

WGan (already tried in GAN repo): https://lilianweng.github.io/lil-log/2017/08/20/from-GAN-to-WGAN.html

https://adeshpande3.github.io/Deep-Learning-Research-Review-Week-1-Generative-Adversarial-Nets

https://machinelearning.apple.com/2017/07/07/GAN.html

EnhanceNet: Single Image Super-Resolution Through Automated Texture Synthesis

[Clorr]

A note on blurry reconstruction for mse: https://wiseodd.github.io/techblog/2017/02/09/why-l2-blurry/

[Clorr]

A slideshare with a lot of different architectures: https://fr.slideshare.net/yuhuang/deep-learning-for-image-denoising-superresolution-27435126