Fusing pairwise modalities for emotion recognition in conversations

Multimodal fusion has the potential to significantly enhance model performance in the domain of Emotion Recognition in Conversations (ERC) by efficiently integrating information from diverse modalities. However, existing methods face challenges as they directly integrate information from different m...

Full description

Saved in:
Bibliographic Details
Main Authors: Fan, Chunxiao, Lin, Jie, Mao, Rui, Cambria, Erik
Other Authors: School of Computer Science and Engineering
Format: Article
Language:English
Published: 2024
Subjects:
Online Access:https://hdl.handle.net/10356/175811
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English