EMO-R3: Reflective Reinforcement Learning for Emotional Reasoning in Multimodal Large Language Models

TL;DR

EMO-R3 introduces a reflective reinforcement learning framework that enhances emotional reasoning and interpretability in multimodal large language models, leading to better performance on visual emotional understanding tasks.

Contribution

The paper proposes a novel Reflective Reinforcement Learning framework with structured emotional thinking and emotional rewards to improve emotional reasoning in MLLMs.

Findings

Significant improvement in emotional reasoning accuracy.

Enhanced interpretability of model decisions.

Better alignment with human emotional understanding.

Abstract

Multimodal Large Language Models (MLLMs) have shown remarkable progress in visual reasoning and understanding tasks but still struggle to capture the complexity and subjectivity of human emotions. Existing approaches based on supervised fine-tuning often suffer from limited generalization and poor interpretability, while reinforcement learning methods such as Group Relative Policy Optimization fail to align with the intrinsic characteristics of emotional cognition. To address these challenges, we propose Reflective Reinforcement Learning for Emotional Reasoning (EMO-R3), a framework designed to enhance the emotional reasoning ability of MLLMs. Specifically, we introduce Structured Emotional Thinking to guide the model to perform step-by-step emotional reasoning in a structured and interpretable manner, and design a Reflective Emotional Reward that enables the model to re-evaluate its reasoning based on visual-text consistency and emotional coherence. Extensive experiments demonstrate that EMO-R3 significantly improves both the interpretability and emotional intelligence of MLLMs, achieving superior performance across multiple visual emotional understanding benchmarks.