情感识别大模型研究综述
Research Survey on Emotion Recognition Large Language Models
摘要
Abstract
The rapid development of large language models offers a new paradigm for emotion recognition,demonstrating significant advantages over conventional methods in complex scenario understanding,zero/few-shot generalization,and multimodal collaborative representation.This paper provides a systematic review of large-scale emotion recognition models.It first summarizes advances in unimodal emotion recognition across text,speech,visual,and physiological signals.The study then focuses on multimodal approaches,categorizing them into unified encoder architectures,hierarchical fusion architectures,and generative model architectures based on fusion strategies,and compares their design principles,key techniques,and applicable scenarios.Further analysis highlights improvements in recognition accuracy,generalization ca-pability,and multimodal fusion,underscoring the important role and potential of large language models in this field.Finally,this paper identifies challenges such as insufficient temporal emotion modeling and cross-cultural cognitive biases,and suggests future directions including enhancing long-range temporal modeling and establishing multimodal cultural emo-tion benchmarks.关键词
单模态情感识别/多模态情感识别/多源信息融合/大模型技术Key words
unimodal emotion recognition/multimodal emotion recognition/multi-source information fusion/large lan-guage model technology分类
信息技术与安全科学引用本文复制引用
吴敖,王海龙,柳林,史文韬..情感识别大模型研究综述[J].计算机科学与探索,2026,20(3):625-649,25.基金项目
国家自然科学基金(62566047) （62566047）
内蒙古自治区自然科学基金(2023LHMS06006,2024LHMS06015) （2023LHMS06006,2024LHMS06015）
内蒙古师范大学基本科研业务费专项资金(2022JBYJ032) （2022JBYJ032）
内蒙古自治区档案馆档案科技项目(2023-13) （2023-13）
无穷维哈密顿系统及其算法应用教育部重点实验室(内蒙古师范大学)开放课题(2023KFYB03) （内蒙古师范大学）
2022年度国家社科基金冷门绝学研究专项学术团队项目(22VJXT008).This work was supported by the National Natural Science Foundation of China(62566047),the Natural Science Foundation of Inner Mongolia(2023LHMS06006,2024LHMS06015),the Fundamental Research Funds for Inner Mongolia Normal University(2022JBYJ032),the Archives Science Project of Inner Mongolia(2023-13),the Project of Key Laboratory of Infinite-Dimensional Ham-iltonian System and Its Algorithm Application(Inner Mongolia Normal University),Ministry of Education(2023KFYB03),and the 2022 Academic Team Project of the National Social Science Fund of China Special Program for Endangered and Obscure Disciplines(22VJXT008). （22VJXT008）
