摘要
文章对现有的基于情感的图像分类研究进行了全面梳理,对图像情感分类过程中的三个关键问题——情感特征提取、情感空间选择和情感识别分类器设计进行了重点分析。同时对现有研究中存在的问题进行了总结,并分析其未来的发展趋势,为后续研究提供了参考。
This paper makes a thorough review of the existing research works in the emotion-based image classification filed. Three key issues： emotional features extraction,emotional space selection and the design of emotion recognition classifier in the process of image emotion classification are analyzed emphatically. At the same time,the paper summarizes the problems in the existing research works and further discusses the future research directions. The paper can give references to the follow-up study.
出处
《情报理论与实践》
CSSCI
北大核心
2018年第6期149-153,160,共6页
Information Studies:Theory & Application
基金
国家自然科学基金项目"基于情感倾向性分析的网络舆情意见领袖识别与对策研究"的成果
项目编号:71303111
关键词
图像情感分类
情感语义
情感特征
情感分析
研究进展
image emotion classification
emotional semantics
emotional features
emotional analysis
research progress
