Timezone: »

Twins: Revisiting the Design of Spatial Attention in Vision Transformers
Xiangxiang Chu · Zhi Tian · Yuqing Wang · Bo Zhang · Haibing Ren · Xiaolin Wei · Huaxia Xia · Chunhua Shen

Thu Dec 09 12:30 AM -- 02:00 AM (PST) @ None #None

Very recently, a variety of vision transformer architectures for dense prediction tasks have been proposed and they show that the design of spatial attention is critical to their success in these tasks. In this work, we revisit the design of the spatial attention and demonstrate that a carefully devised yet simple spatial attention mechanism performs favorably against the state-of-the-art schemes. As a result, we propose two vision transformer architectures, namely, Twins- PCPVT and Twins-SVT. Our proposed architectures are highly efficient and easy to implement, only involving matrix multiplications that are highly optimized in modern deep learning frameworks. More importantly, the proposed architectures achieve excellent performance on a wide range of visual tasks including image-level classification as well as dense detection and segmentation. The simplicity and strong performance suggest that our proposed architectures may serve as stronger backbones for many vision tasks.

Author Information

Xiangxiang Chu (Xiaomi)
Zhi Tian
Yuqing Wang (Nankai University)
Bo Zhang (Meituan)
Haibing Ren (Alibaba Inc)
Xiaolin Wei (Texas A&M)
Huaxia Xia (Meituan)
Chunhua Shen (University of Adelaide)

More from the Same Authors