MCAN splits cross-modal features into aligned and conflicting parts using singular value decomposition and reports the best Acc2, Acc7, and F1 scores on CMU-MOSI and CMU-MOSEI among the compared methods.
Tensor fusion network for mul- timodal sentiment analysis,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
other 1
citation-polarity summary
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1roles
other 1polarities
unclear 1representative citing papers
citing papers explorer
-
Multi-level Conflict-Aware Network for Multi-modal Sentiment Analysis
MCAN splits cross-modal features into aligned and conflicting parts using singular value decomposition and reports the best Acc2, Acc7, and F1 scores on CMU-MOSI and CMU-MOSEI among the compared methods.