Abstract:In the real world, multimodal sentiment analysis (MSA) enables the capture and analysis of sentiments by fusing multimodal information, thereby enhancing the understanding of real-world environments. The key challenges lie in handling the noise in the acquired data and achieving effective multimodal fusion. When processing the noise in data, existing methods utilize the combination of multimodal features to mitigate errors in sentiment word recognition caused by the performance limitations of automatic speech recognition (ASR) models. However, there still remains the problem of how to more efficiently utilize and combine different modalities to address the data noise. In multimodal fusion, most existing fusion methods have limited adaptability to the feature differences between modalities, making it difficult to capture the potential complex nonlinear interactions that may exist between modalities. To overcome the aforementioned issues, this paper proposes a new framework named multimodal-word-refinement and cross-modal-hierarchy (MWRCMH) fusion. Specifically, we utilized a multimodal word correction module to reduce sentiment word recognition errors caused by ASR. During multimodal fusion, we designed a cross-modal hierarchical fusion module that employed cross-modal attention mechanisms to fuse features between pairs of modalities, resulting in fused bimodal-feature information. Then, the obtained bimodal information and the unimodal information were fused through the nonlinear layer to obtain the final multimodal sentiment feature information. Experimental results on the MOSI-SpeechBrain, MOSI-IBM, and MOSI-iFlytek datasets demonstrated that the proposed approach outperformed other comparative methods, achieving Has0-F1 scores of 76.43%, 80.15%, and 81.93%, respectively. Our approach exhibited better performance, as compared to multiple baselines.

A Relation-Aware Heterogeneous Graph Transformer on Dynamic Fusion for Multimodal Classification Tasks

Sentiment Analysis Using Deep Robust Complementary Fusion of Multi-Features and Multi-Modalities.

Multimodal Sentiment Analysis Based on Cross-Modal Attention and Gated Cyclic Hierarchical Fusion Networks

Heterogeneous Hierarchical Fusion Network for Multimodal Sentiment Analysis in Real-World Environments

Tri-Modalities Fusion for Multimodal Sentiment Analysis

Multimodal Sentiment Analysis Using Multi-tensor Fusion Network with Cross-modal Modeling

A cross modal hierarchical fusion multimodal sentiment analysis method based on multi-task learning

Multi-scale Cooperative Multimodal Transformers for Multimodal Sentiment Analysis in Videos

Deep Symmetric Fusion Transformer for Multimodal Remote Sensing Data Classification

Hybrid Transformer with Multi-level Fusion for Multimodal Knowledge Graph Completion

DGFN Multimodal Emotion Analysis Model Based on Dynamic Graph Fusion Network

Cross-Modal Sentiment Sensing with Visual-Augmented Representation and Diverse Decision Fusion

MATF: main-auxiliary transformer fusion for multi-modal sentiment analysis

A transformer-encoder-based multimodal multi-attention fusion network for sentiment analysis

Multimodal feature fusion based on object relation for video captioning

Multimodal Sentiment Analysis in Realistic Environments Based on Cross-Modal Hierarchical Fusion Network

Mutually Beneficial Transformer for Multimodal Data Fusion

Fusion with Hierarchical Graphs for Mulitmodal Emotion Recognition

Multi-Channel Attentive Graph Convolutional Network with Sentiment Fusion for Multimodal Sentiment Analysis

Multi-Graph Heterogeneous Interaction Fusion for Social Recommendation

Multimodal Remote Sensing Data Classification Based on Gaussian Mixture Variational Dynamic Fusion Network