Pixel Recursive Super Resolution
This addresses the issue of blurry edges in image super-resolution for applications like photo enhancement, though it is incremental as it builds on existing PixelCNN architectures.
The paper tackles the problem of generating realistic high-resolution images from low-resolution inputs by modeling the super-resolution process as a multimodal conditional distribution, and results show that samples from the model are rated as more photo-realistic than an L2 regression baseline in human evaluations.
We present a pixel recursive super resolution model that synthesizes realistic details into images while enhancing their resolution. A low resolution image may correspond to multiple plausible high resolution images, thus modeling the super resolution process with a pixel independent conditional model often results in averaging different details--hence blurry edges. By contrast, our model is able to represent a multimodal conditional distribution by properly modeling the statistical dependencies among the high resolution image pixels, conditioned on a low resolution input. We employ a PixelCNN architecture to define a strong prior over natural images and jointly optimize this prior with a deep conditioning convolutional network. Human evaluations indicate that samples from our proposed model look more photo realistic than a strong L2 regression baseline.