StarSRGAN: Improving Real-World Blind Super-Resolution

TL;DR

StarSRGAN introduces a new GAN-based model for blind super-resolution that outperforms previous methods in quality and speed, enabling real-time high-resolution image reconstruction.

Contribution

The paper presents StarSRGAN, a novel GAN model with five architectures that achieves state-of-the-art performance and includes a lightweight version for real-time applications.

Findings

Approximately 10% better on MANIQA and AHIQ measures compared to Real-ESRGAN.

StarSRGAN Lite is about 7.5 times faster, enabling real-time 4K upsampling.

Maintains nearly 90% of image quality with significantly reduced speed.

Abstract

The aim of blind super-resolution (SR) in computer vision is to improve the resolution of an image without prior knowledge of the degradation process that caused the image to be low-resolution. The State of the Art (SOTA) model Real-ESRGAN has advanced perceptual loss and produced visually compelling outcomes using more complex degradation models to simulate real-world degradations. However, there is still room to improve the super-resolved quality of Real-ESRGAN by implementing recent techniques. This research paper introduces StarSRGAN, a novel GAN model designed for blind super-resolution tasks that utilize 5 various architectures. Our model provides new SOTA performance with roughly 10% better on the MANIQA and AHIQ measures, as demonstrated by experimental comparisons with Real-ESRGAN. In addition, as a compact version, StarSRGAN Lite provides approximately 7.5 times faster reconstruction speed (real-time upsampling from 540p to 4K) but can still keep nearly 90% of image quality, thereby facilitating the development of a real-time SR experience for future research. Our codes are released at https://github.com/kynthesis/StarSRGAN.