WaveMixSR-V2: Enhancing Super-resolution with Higher Efficiency

TL;DR

WaveMixSR-V2 advances single image super-resolution by integrating a pixel shuffle and multistage design, achieving state-of-the-art results with improved efficiency and resource utilization.

Contribution

The paper introduces WaveMixSR-V2, a novel super-resolution architecture that enhances performance and efficiency over previous models by replacing transpose convolution with pixel shuffle and adding a multistage design.

Findings

Outperforms existing architectures on BSD100 dataset

Achieves higher parameter efficiency and lower latency

Demonstrates superior resource utilization and throughput

Abstract

Recent advancements in single image super-resolution have been predominantly driven by token mixers and transformer architectures. WaveMixSR utilized the WaveMix architecture, employing a two-dimensional discrete wavelet transform for spatial token mixing, achieving superior performance in super-resolution tasks with remarkable resource efficiency. In this work, we present an enhanced version of the WaveMixSR architecture by (1) replacing the traditional transpose convolution layer with a pixel shuffle operation and (2) implementing a multistage design for higher resolution tasks ($4\times$). Our experiments demonstrate that our enhanced model -- WaveMixSR-V2 -- outperforms other architectures in multiple super-resolution tasks, achieving state-of-the-art for the BSD100 dataset, while also consuming fewer resources, exhibits higher parameter efficiency, lower latency and higher throughput. Our code is available at https://github.com/pranavphoenix/WaveMixSR.