DualFormer: Local-global stratified transformer for efficient video recognition

• Main Authors: LIANG, Yuxuan, ZHOU, Pan, ZIMMERMANN, Roger, YAN, Shuicheng
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2022
• Subjects: Efficient video transformer; Local and global attention; Artificial Intelligence and Robotics; Graphics and Human Computer Interfaces
• Online Access: https://ink.library.smu.edu.sg/sis_research/8980; https://ink.library.smu.edu.sg/context/sis_research/article/9983/viewcontent/2022_ECCV_DualFormer.pdf
• Institution: Singapore Management University

While transformers have shown great potential on video recognition with their strong capability of capturing long-range dependencies, they often suffer high computational costs induced by the self-attention to the huge number of 3D tokens. In this paper, we present a new transformer architecture ter...