A Review of Multi-modal Human Motion Recognition Based on Deep Learning

作     者：Ye Li Yifan Pan Xinhui Wu 

作者机构：Software CollegeShenyang Normal University Shenyang 110034 China 

出 版 物：《IJLAI Transactions on Science and Engineering》 (IJLAI科学与工程学报汇刊（英文）)

年 卷 期：2024年第2卷第3期

页      面：36-46页

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

基　　金：2021 Scientific research funding project of Liaoning Provincial Education Department(Research and implementation of university scientific research information platform serving the transformation of achievements) 

主　　题：Human motion recognition Computer vision Multi-modal Deep learning 

摘      要：Human motion recognition is a research hotspot in the field of computer vision,which has a wide range of applications,including biometrics,intelligent surveillance and human-computer *** visionbased human motion recognition,the main input modes are RGB,depth image and bone *** mode can capture some kind of information,which is likely to be complementary to other modes,for example,some modes capture global information while others capture local details of an *** speaking,the fusion of multiple modal data can improve the recognition *** addition,how to correctly model and utilize spatiotemporal information is one of the challenges facing human motion *** at the feature extraction methods involved in human action recognition tasks in video,this paper summarizes the traditional manual feature extraction methods from the aspects of global feature extraction and local feature extraction,and introduces the commonly used feature learning models of feature extraction methods based on deep learning in *** paper summarizes the opportunities and challenges in the field of motion recognition and looks forward to the possible research directions in the future.