arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2204.07366cs.CV
收藏

ResT V2:更简单、更快速、更强大

ResT V2: Simpler, Faster and Stronger

State Key Laboratory for Novel Software Technology(新型软件技术国家重点实验室);Nanjing University(南京大学)

Qing-Long Zhang, Yu-Bin Yang

摘要:

本文提出了ResTv2,一个更简单,更快,更强大的多尺度视觉Transformer的视觉识别。ResTv2简化了ResTv1中的EMSA结构(即,消除多头交互部分),并采用上采样操作来重建由下采样操作引起的丢失的中频和高频信息。此外,我们还探索了不同的技术,以更好地将ResTv2骨干应用于下游任务。我们发现,虽然结合EMSAv2和窗口注意力可以大大降低理论矩阵乘法FLOP,但它可能会显着降低计算密度,从而导致实际速度降低。我们在ImageNet分类、COCO检测和ADE 20 K语义分割上全面验证了ResTv2。实验结果表明,所提出的ResTv2可以大大优于最近的国家的最先进的骨干,证明了ResTv2作为坚实的骨干的潜力。代码和模型将在\url{https://github.com/wofmanaf/ResT}上公开发布

英文摘要:

This paper proposes ResTv2, a simpler, faster, and stronger multi-scale vision Transformer for visual recognition. ResTv2 simplifies the EMSA structure in ResTv1 (i.e., eliminating the multi-head interaction part) and employs an upsample operation to reconstruct the lost medium- and high-frequency information caused by the downsampling operation. In addition, we explore different techniques for better apply ResTv2 backbones to downstream tasks. We found that although combining EMSAv2 and window attention can greatly reduce the theoretical matrix multiply FLOPs, it may significantly decrease the computation density, thus causing lower actual speed. We comprehensively validate ResTv2 on ImageNet classification, COCO detection, and ADE20K semantic segmentation. Experimental results show that the proposed ResTv2 can outperform the recently state-of-the-art backbones by a large margin, demonstrating the potential of ResTv2 as solid backbones. The code and models will be made publicly available at \url{https://github.com/wofmanaf/ResT}

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。

↑