Leveraging Self-Paced Curriculum Learning for Enhanced Modality Balance in Multimodal Conversational Emotion Recognition

TL;DR

This paper introduces a Self-Paced Curriculum Learning framework to improve modality balance and robustness in Multimodal Emotion Recognition in Conversations, demonstrating consistent performance gains across datasets.

Contribution

It proposes a dual-level Difficulty Measurer and Learning Scheduler to dynamically guide training, addressing modality imbalance in MERC.

Findings

SPCL improves weighted F1-score by up to +6.6% on IEMOCAP.

SPCL achieves up to +10.4% gains on MELD.

The method enhances robustness across different architectures.

Abstract

Multimodal Emotion Recognition in Conversations (MERC) is a crucial task for understanding human interactions, where multimodal approaches integrating language, facial expressions, and vocal tone have achieved significant progress. However, modality misalignment and imbalanced learning remain major challenges, limiting the effective utilization of multimodal information. To address this issue, we propose a plug-and-play framework based on Self-Paced Curriculum Learning (SPCL) for MERC. We introduce a dual-level Difficulty Measurer that captures both utterance-level and conversation-level challenges. The utterance-level score models fine-grained modality-specific difficulty, while the conversation-level score captures broader dialogue structures, including emotional dependencies and modality coherence. Based on these scores, the Learning Scheduler dynamically guides training from easier to more difficult instances. By integrating SPCL into existing MERC architectures, our method alleviates modality imbalance and improves model robustness. Extensive experiments on the IEMOCAP and MELD datasets demonstrate consistent improvements across different architectures and modality settings. On IEMOCAP, SPCL improves weighted F1-score by approximately +1.2% to +6.6% over baseline models, while on MELD, gains reach up to +10.4%. These results highlight the effectiveness and generalizability of SPCL as a lightweight plug-and-play module for multimodal emotion recognition.