arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2012.08510cs.CV
收藏

GTA:用于视频动作理解的全局时间注意力

GTA: Global Temporal Attention for Video Action Understanding

University of Maryland(马里兰大学);Fudan University(复旦大学);Facebook AI(Facebook人工智能)

Bo He, Xitong Yang, Zuxuan Wu, Hao Chen, Ser-Nam Lim, Abhinav Shrivastava

摘要:

自我注意力学习成对交互来建模长距离依赖关系,从而极大地改进了视频动作识别。在本文中,我们寻求更深入的理解自注意视频中的时间建模。我们首先证明,纠缠模型的时空信息平坦的所有像素是次优的,无法捕捉帧之间的时间关系明确。为此,我们引入了全局时间注意力(GTA),它以解耦的方式在空间注意力之上执行全局时间注意力。我们将GTA应用于像素和语义相似的区域,以捕获不同级别的空间粒度的时间关系。与传统的计算特定于实例的注意力矩阵的自注意力不同,GTA直接学习全局注意力矩阵,该矩阵旨在对跨不同样本进行概括的时间结构进行编码。我们进一步增强GTA与跨通道多头时尚,利用更好的时间建模通道的相互作用。在2D和3D网络上进行的大量实验表明,我们的方法始终增强了时间建模,并在三个视频动作识别数据集上提供了最先进的性能。

英文摘要:

Self-attention learns pairwise interactions to model long-range dependencies, yielding great improvements for video action recognition. In this paper, we seek a deeper understanding of self-attention for temporal modeling in videos. We first demonstrate that the entangled modeling of spatio-temporal information by flattening all pixels is sub-optimal, failing to capture temporal relationships among frames explicitly. To this end, we introduce Global Temporal Attention (GTA), which performs global temporal attention on top of spatial attention in a decoupled manner. We apply GTA on both pixels and semantically similar regions to capture temporal relationships at different levels of spatial granularity. Unlike conventional self-attention that computes an instance-specific attention matrix, GTA directly learns a global attention matrix that is intended to encode temporal structures that generalize across different samples. We further augment GTA with a cross-channel multi-head fashion to exploit channel interactions for better temporal modeling. Extensive experiments on 2D and 3D networks demonstrate that our approach consistently enhances temporal modeling and provides state-of-the-art performance on three video action recognition datasets.

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。

↑