Understanding multimodal sentiment with deep modality interaction learning

Jie Mu

doi:10.1016/j.patcog.2026.113236

Understanding multimodal sentiment with deep modality interaction learning

2026-02-05

0

PRE

AI

DOI：10.1016/j.patcog.2026.113236

Original

Original request for help

Save

Abstract

En 中文

• Overcome the limitations of existing multimodal sentiment analysis methods. • Construct a cross-modal graph to jointly model intra- and inter-modal interactions. • Enhance visual representations using text-guided interactive attention mechanisms. • Leverage graph convolution to aggregate complementary multimodal sentiment cues. • Quantitative and qualitative experiment results of proposed method are excellent.

Keywords:

multimodal sentiment analysis

cross-modal graph

interactive attention

graph convolution

modality interaction

Understanding multimodal sentiment with deep modality interaction learning

Abstract

Journal

Organization