Large Kernel Distillation Network for Efficient Single Image Super-Resolution

TL;DR

This paper introduces the Large Kernel Distillation Network (LKDN), a lightweight super-resolution model that uses large kernels, reparameterization, and novel optimization to outperform existing methods with reduced computational costs.

Contribution

The paper proposes LKDN, a novel lightweight SISR model that simplifies structure, employs efficient attention, and introduces a new optimizer to enhance performance and training speed.

Findings

LKDN outperforms existing lightweight SR methods.

LKDN achieves state-of-the-art performance.

The model reduces computational costs while maintaining high accuracy.

Abstract

Efficient and lightweight single-image super-resolution (SISR) has achieved remarkable performance in recent years. One effective approach is the use of large kernel designs, which have been shown to improve the performance of SISR models while reducing their computational requirements. However, current state-of-the-art (SOTA) models still face problems such as high computational costs. To address these issues, we propose the Large Kernel Distillation Network (LKDN) in this paper. Our approach simplifies the model structure and introduces more efficient attention modules to reduce computational costs while also improving performance. Specifically, we employ the reparameterization technique to enhance model performance without adding extra cost. We also introduce a new optimizer from other tasks to SISR, which improves training speed and performance. Our experimental results demonstrate that LKDN outperforms existing lightweight SR methods and achieves SOTA performance.