Efficient Mixed Transformer for Single Image Super-Resolution
Authors:
Preprint
Abstract
Recently, Transformer-based methods have achieved impressive results in single image super-resolution (SISR). However, the lack of locality mechanism and high complexity limit their application in the field of super-resolution (SR). To solve these problems, we propose a new method, Efficient Mixed Transformer (EMT) in this study. Specifically, we propose the Mixed Transformer Block (MTB), consisting of multiple consecutive transformer layers, in some of which the Pixel Mixer (PM) is used to replace the Self-Attention (SA). PM can enhance the local knowledge aggregation with pixel shifting operations. At the same time, no additional complexity is introduced as PM has no parameters and floating-point operations. Moreover, we employ striped window for SA (SWSA) to gain an efficient global dependency modelling by utilizing image anisotropy. Experimental results show that EMT outperforms the existing methods on benchmark dataset and achieved state-of-the-art performance. The Code is available at https://github. com/Fried-Rice-Lab/EMT.git.Keywords
All Related Versions
This publication has 5 references indexed in Scilit:
- On Efficient Transformer-Based Image Pre-training for Low-Level Vision2023
- N-Gram in Swin Transformers for Efficient Lightweight Image Super-Resolution2023
- Efficient and Explicit Modelling of Image Hierarchies for Image Restoration2023
- Efficient Long-Range Attention Network for Image Super-Resolution2022
- Shift: A Zero FLOP, Zero Parameter Alternative to Spatial Convolutions2018