Modal Interactive Feature Encoder for Multimodal Sentiment Analysis

作     者：Xiaowei Zhao Jie Zhou Xiujuan Xu 

作者机构：School of SoftwareDalian University of TechnologyDalian 116620China Key Laboratory for Ubiquitous Network and Service Software of Liaoning ProvinceDalian 116620China 

出 版 物：《国际计算机前沿大会会议论文集》 (International Conference of Pioneering Computer Scientists, Engineers and Educators（ICPCSEE）)

年 卷 期：2023年第2期

页      面：285-303页

学科分类：0821[工学-纺织科学与工程] 08[工学] 

主　　题：Multimodal Sentiment Analysis Modal Interaction Feature Encoder 

摘      要：Multimodal Sentiment analysis refers to analyzing emotions in infor-mation carriers containing multiple *** better analyze the features within and between modalities and solve the problem of incomplete multimodal feature fusion,this paper proposes a multimodal sentiment analysis model MIF(Modal Interactive Feature Encoder For Multimodal Sentiment Analysis).First,the global features of three modalities are obtained through unimodal feature extraction ***,the inter-modal interactive feature encoder and the intra-modal interactive feature encoder extract similarity features between modal-ities and intra-modal special features ***,unimodal special features and the interaction information between modalities are decoded to get the fusion features and predict sentimental polarity *** conduct extensive experi-ments on three public multimodal datasets,including one in Chinese and two in *** results show that the performance of our approach is significantly improved compared with benchmark models.