Efficient Super Resolution For Large-Scale Images Using Attentional GAN
This addresses a practical need for large-scale image super resolution in business applications like online displays, but it is incremental as it adapts existing GAN methods to a new scale.
The paper tackles the problem of generating high-resolution images (at least 2000px) for commercial use, where existing methods focus on smaller sizes, and reports that their attentional GAN model achieves high PSNR and SSIM scores with a training speedup of around five times.
Single Image Super Resolution (SISR) is a well-researched problem with broad commercial relevance. However, most of the SISR literature focuses on small-size images under 500px, whereas business needs can mandate the generation of very high resolution images. At Expedia Group, we were tasked with generating images of at least 2000px for display on the website, four times greater than the sizes typically reported in the literature. This requirement poses a challenge that state-of-the-art models, validated on small images, have not been proven to handle. In this paper, we investigate solutions to the problem of generating high-quality images for large-scale super resolution in a commercial setting. We find that training a generative adversarial network (GAN) with attention from scratch using a large-scale lodging image data set generates images with high PSNR and SSIM scores. We describe a novel attentional SISR model for large-scale images, A-SRGAN, that uses a Flexible Self Attention layer to enable processing of large-scale images. We also describe a distributed algorithm which speeds up training by around a factor of five.