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 2107.00641cs.CVcs.AIcs.LG
收藏

视觉Transformer中局部-全局相互作用的焦点自注意

Focal Self-attention for Local-Global Interactions in Vision Transformers

Microsoft Research(微软研究院);Microsoft Cloud + AI(微软云计算与人工智能部门)

Jianwei Yang, Chunyuan Li, Pengchuan Zhang, Xiyang Dai, Bin Xiao, Lu Yuan, Jianfeng Gao

摘要:

最近,视觉Transformer及其变体在各种计算机视觉任务上显示出巨大的前景。通过自我注意捕捉短期和长期视觉依赖的能力可以说是成功的主要来源。但由于二次计算开销,它也带来了挑战,特别是对于高分辨率视觉任务(例如,目标检测)。在本文中,我们提出了焦点自关注,一个新的机制,结合了细粒度的本地和粗粒度的全球相互作用。使用这种新的机制,每个令牌出席最近的周围令牌在细粒度,但令牌远离在粗粒度,因此可以捕获短期和长期的视觉依赖有效和有效的。焦点自我关注,我们提出了一个新的变体的视觉Transformer模型,称为焦点Transformer,它实现了卓越的性能超过国家的最先进的视觉Transformer的一系列公共图像分类和目标检测基准。特别是,我们的Focal Transformer模型具有中等大小的51.1M和更大的89.8M,在224 x224分辨率的ImageNet分类上分别实现了83.5和83.8的Top-1准确度。使用焦点Transformers作为骨干,我们获得了一致的和实质性的改进,超过目前最先进的Swin Transformers的6种不同的目标检测方法训练与标准的1x和3x时间表。我们最大的Focal Transformer在COCO mini-val/test-dev上产生58.7/58.9的框mAP和50.9/51.3的掩码mAP,在ADE 20 K上产生55.4 mIoU的语义分割,在三个最具挑战性的计算机视觉任务上创建新的SoTA。

英文摘要:

Recently, Vision Transformer and its variants have shown great promise on various computer vision tasks. The ability of capturing short- and long-range visual dependencies through self-attention is arguably the main source for the success. But it also brings challenges due to quadratic computational overhead, especially for the high-resolution vision tasks (e.g., object detection). In this paper, we present focal self-attention, a new mechanism that incorporates both fine-grained local and coarse-grained global interactions. Using this new mechanism, each token attends the closest surrounding tokens at fine granularity but the tokens far away at coarse granularity, and thus can capture both short- and long-range visual dependencies efficiently and effectively. With focal self-attention, we propose a new variant of Vision Transformer models, called Focal Transformer, which achieves superior performance over the state-of-the-art vision Transformers on a range of public image classification and object detection benchmarks. In particular, our Focal Transformer models with a moderate size of 51.1M and a larger size of 89.8M achieve 83.5 and 83.8 Top-1 accuracy, respectively, on ImageNet classification at 224x224 resolution. Using Focal Transformers as the backbones, we obtain consistent and substantial improvements over the current state-of-the-art Swin Transformers for 6 different object detection methods trained with standard 1x and 3x schedules. Our largest Focal Transformer yields 58.7/58.9 box mAPs and 50.9/51.3 mask mAPs on COCO mini-val/test-dev, and 55.4 mIoU on ADE20K for semantic segmentation, creating new SoTA on three of the most challenging computer vision tasks.

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。

↑