Motion Enhanced Model Based on High-Level Spatial Features

作     者：Yang Wu Lei Guo Xiaodong Dai Bin Zhang Dong-Won Park Ming Ma 

作者机构：College of Computer Science and EngineeringInner Mongolia UniversityHohhot010021China Department of Information and CommunicationsPaiChai UniversityDaejeon35345Korea 

出 版 物：《Computers, Materials & Continua》 (计算机、材料和连续体（英文）)

年 卷 期：2022年第73卷第12期

页      面：5911-5924页

核心收录：

学科分类：0710[理学-生物学] 08[工学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：This work is supported by the Inner Mongolia Natural Science Foundation of China under Grant No.2021MS06016 the CERNET Innovation Project(NGII20190625) 

主　　题：Action recognition motion vectors two-stream knowledge distillation accumulate strategy 

摘      要：Action recognition has become a current research hotspot in computer *** to other deep learning methods,Two-stream convolutional network structure achieves better performance in action recognition,which divides the network into spatial and temporal streams,using video frame images as well as dense optical streams in the network,respectively,to obtain the category ***,the two-stream network has some drawbacks,i.e.,using dense optical flow as the input of the temporal stream,which is computationally expensive and extremely time-consuming for the current extraction algorithm and cannot meet the requirements of real-time *** this paper,instead of the dense optical flow,the Motion Vectors(MVs)are used and extracted from the compressed domain as temporal features,which greatly reduces the extraction ***,the motion pattern that MVs contain is coarser,which leads to low *** this paper,we propose two strategies to improve the accuracy:firstly,an accumulated strategy is used to enhance the motion information and continuity of MVs;secondly,knowledge distillation is used to fuse the spatial information into the temporal stream so that more information(e.g.,motion details,colors,etc.)is *** results show that the accuracy of MV can be greatly improved by the strategies proposed in this paper and the final recognition for human actions accuracy is guaranteed without using optical flow.