摘要
情感计算是现代人机交互中的一个重要研究方向,旨在研究与开发能够识别、解释、处理和模拟人类情感的理论、方法与系统.脑电、心电、皮肤电等生理信号是情感计算中重要的输入信号.本文总结了近年来基于脑电等生理信号的情感计算研究所取得的进展.首先介绍情感计算的相关基础理论,不同生理信号与情感变化之间的联系,以及基于生理信号的情感计算工作流程和相关公开数据集.接下来介绍生理信号的特征工程和情感计算中的机器学习算法,重点介绍适合处理个体差异的迁移学习、降低数据标注量的主动学习和融合特征工程与学习器的深度学习算法.最后,指出基于生理信号的情感计算研究中面临的一些挑战.
Affective computing is an important research area in modern human-machine interaction.It aims to develop systems that can recognize,interpret,analyze and emulate human emotions.Physiological signals,e.g.,electroencephalogram,electrocardiogram,galvanic skin response,etc.,are important inputs in affective computing.This paper summarizes recent progresses on physiological signals based affective computing,particularly,electroencephalogram based affective computing.It first introduces the basic theories of affective computing,the relationship between the changes of physiological signals and affects,the flowchart of physiological signals based affective computing,and common public datasets.Next,it introduces typical feature engineering and machine learning algorithms,particularly transfer learning,active learning and deep learning.Finally,it points out some challenges and future research directions in physiological signals based affective computing.
作者
权学良
曾志刚
蒋建华
张亚倩
吕宝粮
伍冬睿
QUAN Xue-Liang;ZENG Zhi-Gang;JIANG Jian-Hua;ZHANG Ya-Qian;LV Bao-Liang;WU Dong-Rui(Ministry of Education Key Laboratory on Image Information Processing and Intelligent Control,School of Artificial Intelligence and Automation,Huazhong University of Science and Technology,Wuhan 430074;Department of Computer Science and Engineering,Shanghai Jiao Tong University,Shanghai 200240)
出处
《自动化学报》
EI
CAS
CSCD
北大核心
2021年第8期1769-1784,共16页
Acta Automatica Sinica
基金
湖北省技术创新专项资助项目(2019AEA171)
湖北省杰出青年基金(2020CFA050)
武汉市应用基础前沿项目(2020020601012240)
国家自然科学基金(61673266,61976135)资助。
关键词
情感计算
情绪分类
脑机接口
机器学习
Affective computing
emotion classification
brain-computer interface
machine learning
