arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2409.03516cs.CVcs.AI
收藏

LMLT:用于图像超分辨率的低到高多级视觉Transformer

LMLT: Low-to-high Multi-Level Vision Transformer for Image Super-Resolution

Sogang University(西江大学)

Jeongsoo Kim, Jongho Nang, Junsuk Choe

AI总结:

本文提出LMLT,通过沿通道维度划分特征并为不同头设置递减空间尺寸的自注意力,在降低推理时间和GPU内存的同时保持或超越现有ViT超分辨率方法性能。

摘要:

近年来,基于视觉Transformer(ViT)的图像超分辨率方法已展现出令人瞩目的性能。然而,它们存在显著的复杂度问题,导致推理时间长且内存占用高。此外,采用窗口自注意力(WSA)的ViT模型在处理窗口之外的区域时面临挑战。为解决这些问题,我们提出了低到高多级Transformer(LMLT),该方法为每个头使用不同特征尺寸的注意力。LMLT沿通道维度划分图像特征,逐步降低较低头的空间尺寸,并对每个头应用自注意力。该方法能够有效捕获局部和全局信息。通过将较低头的结果整合到较高头中,LMLT克服了自注意力中的窗口边界问题。大量实验表明,我们的模型显著降低了推理时间和GPU内存占用,同时保持甚至超越了最先进的基于ViT的图像超分辨率方法的性能。我们的代码可在https://github.com/jwgdmkj/LMLT获取。

英文摘要:

Recent Vision Transformer (ViT)-based methods for Image Super-Resolution have demonstrated impressive performance. However, they suffer from significant complexity, resulting in high inference times and memory usage. Additionally, ViT models using Window Self-Attention (WSA) face challenges in processing regions outside their windows. To address these issues, we propose the Low-to-high Multi-Level Transformer (LMLT), which employs attention with varying feature sizes for each head. LMLT divides image features along the channel dimension, gradually reduces spatial size for lower heads, and applies self-attention to each head. This approach effectively captures both local and global information. By integrating the results from lower heads into higher heads, LMLT overcomes the window boundary issues in self-attention. Extensive experiments show that our model significantly reduces inference time and GPU memory usage while maintaining or even surpassing the performance of state-of-the-art ViT-based Image Super-Resolution methods. Our codes are availiable at https://github.com/jwgdmkj/LMLT.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑