MELD

The MELD dataset is a benchmark used for evaluating multimodal emotion recognition in conversations, containing annotated utterances that include textual, acoustic, and visual cues.

Papers using MELD (4)

Disentangled Dual-Branch Graph Learning for Conversational Emotion Recognition2026Effective Context Modeling Framework for Emotion Recognition in Conversations2024 · 1 citesLineConGraphs: Line Conversation Graphs for Effective Emotion Recognition using Graph Neural Networks2023Long-Short Distance Graph Neural Networks and Improved Curriculum Learning for Emotion Recognition in Conversation2025