MELD
Emerging3papers using it
2025first seen
The MELD dataset is a multimodal benchmark that contains data for evaluating discrete emotion recognition and continuous sentiment analysis by integrating text, audio, and visual modalities.
Papers using MELD (3)
- A Unified Framework for Emotion Recognition and Sentiment Analysis via Expert-Guided Multimodal Fusion with Large Language ModelsCross-Space Synergy: A Unified Framework for Multimodal Emotion Recognition in ConversationQuality-controlled Multimodal Emotion Recognition In Conversations With Identity-based Transfer Learning And MAMBA Fusion