A new MLLM and dataset combining five affect tasks with a multi-stage instruction-tuning strategy yields strong results on sentiment and emotion benchmarks, but the empirical setup has unresolved comparison and data-release issues.
Semi-supervised mul- timodal emotion recognition with expression mae
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
EmoVerse: Exploring Multimodal Large Language Models for Sentiment and Emotion Understanding
A new MLLM and dataset combining five affect tasks with a multi-stage instruction-tuning strategy yields strong results on sentiment and emotion benchmarks, but the empirical setup has unresolved comparison and data-release issues.