Multimodal Feature Fusion Method for Unbalanced Sample Data in Social Network Public Opinion

Multimodal Feature Fusion Method for Unbalanced Sample Data in Social Network Public Opinion

With the wide application of social media, public opinion analysis in social networks has been unable to be met through text alone because the existing public opinion information includes data information of various modalities, such as voice, text, and facial expressions. Therefore multi-modal emoti...

Full description

Bibliographic Details
Main Authors: Jian Zhao, Wenhua Dong, Lijuan Shi, Wenqian Qiang, Zhejun Kuang, Dawei Xu, Tianbo An
Format: Article
Language:English
Published: MDPI AG 2022-07-01
Series:Sensors
Subjects:
multi-modality
fine-grained
model fusion
emotion features
feature generation
Online Access:https://www.mdpi.com/1424-8220/22/15/5528

Internet

https://www.mdpi.com/1424-8220/22/15/5528

Similar Items