基于时间和运动增强的视频动作识别  

Temporal and motion enhancement for video action recognition

作　　者：汪豪 赵林彬 刘国华[1,2,3,4] WANG Hao;ZHAO Bin;LIU Guo-hua(College of Electronic Information Technology and Optical Engineering,Nankai University,Tianjin 300350,China;Tianjin Key Laboratory of Optoelectronic Sensor and Sensing Network Technology,Tianjin 300350,China;General Terminal IC Interdisciplinary Science Center,Nankai University,Tianjin 300350,China;Engineering Research Center of Thin Film Optoelectronics Technology,Ministry of Education,Nankai University,Tianjin 300350,China;School of Artificial Intelligence,Guilin University of Electronic Technology,Guilin 541004,China)

机构地区：[1]南开大学电子信息与光学工程学院,天津300350 [2]天津市光电传感器与传感网络技术重点实验室,天津300350 [3]南开大学泛终端芯片交叉科学中心,天津300350 [4]南开大学薄膜光电子技术教育部工程研究中心,天津300350 [5]桂林电子科技大学人工智能学院,广西桂林541004

出　　处：《吉林大学学报(工学版)》2025年第1期339-346,共8页Journal of Jilin University:Engineering and Technology Edition

基　　金：中央高校基本科研业务费专项项目。

摘　　要：三维卷积神经网络可以通过直接融合空间和时间特征来实现良好的性能,但计算是密集的。传统的二维卷积神经网络在图像识别中表现良好,但由于无法提取时间特征,导致其在视频动作识别中表现不佳。为此,本文提出一个即插即用的时间和运动增强模块以学习视频动作的时空关系,可以插入任意二维卷积神经网络中,而额外的计算成本是相当有限的。在多个公开动作识别数据集上进行的大量实验表明,本文提出的网络以高效率优于最先进的二维卷积神经网络方法。3D convolutional neural networks can achieve good performance by directly fusing spatial and temporal features but are computationally intensive.Conventional 2D convolutional neural networks perform well in image recognition,but their inability to extract temporal features leads to poor performance in video action recognition.To this end,a plug-and-play temporal and motion enhancement module is proposed to learn spatiotemporal relationships for video action recognition and can be inserted into 2D convolutional neural networks with limited extra computational cost.The extensive experiments on several action recognition datasets demonstrate that the proposed network outperforms the state-of-the-art 2D convolutional neural networks with high efficiency.

关 键 词：计算机应用 动作识别 卷积神经网络 时间建模 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]