Multimodal Hidden Markov Models for Persistent Emotional State Tracking

Anamika Ragu, Aneesh Jonelagadda

arXiv:2605.12838 · 2026-05-14 공개 · arXiv · PDF

llm-as-a-judge dialogue-state-tracking hidden-markov-models multimodal-emotion valence-arousal affective-regimes clinical-conversation context-augmentation

Abstract

Tracking an interpretable emotional arc of a conversation via the sentiment of individual utterances processed as a whole is central to both understanding and guiding communication in applied, especially clinical, conversational contexts. Existing approaches to emotion recognition operate at the utterance level, obscuring the persistent phases that characterize real conversational dynamics. We propose a lightweight framework that models conversational emotion as a sequence of latent emotional regimes using sticky factorial HDP-HMMs over multimodal valence-arousal representations derived from simultaneous video, audio and textual input. We evaluate the quality of regime prediction using LLM-as-a-Judge, geometric, and temporal consistency metrics, demonstrating that the sticky HDP-HMM produces more interpretable regime sequences than the baseline Gaussian HMM at a fraction of the computational cost of LLM-based dialogue state tracking methods. In addition, Question-Answer experiments in a clinical dataset suggest that meaningful emotional phases can reliably be recovered from multimodal valence-arousal trajectories and used to improve the quality of LLM responses in unstable affective regimes via context augmentation. This framework thus opens a path toward interpretable, lightweight, and actionable analysis of conversational emotion dynamics at scale.

한국어 요약

📋 한 줄 요약

**[감정 인식 / HMM]** 멀티모달 valence-arousal 표현 위에 sticky factorial HDP-HMM을 적용해 대화의 지속적 감정 regime을 추적하는 경량 프레임워크 제안.

🎯 핵심 기여도

💡 핵심 아이디어

대화 감정은 발화 단위 점 sequence가 아니라 지속하는 잠재 regime의 sequence다. sticky factorial HDP-HMM은 regime 지속성과 다요인 표현을 동시에 다루어 해석 가능한 감정 호(arc)를 회복한다.

🔬 기술적 접근법

📊 주요 결과

💭 의의 및 한계

**의의**: 임상 대화 보조 같은 안전 핵심 컨텍스트에서 가볍고 해석 가능한 감정 추적 인프라를 제공. **한계**: HDP-HMM 가정에 따른 표현력 한계, 임상 데이터 규모와 도메인 종속성으로 광범위 일반화는 후속 과제.

🚀 실용적 활용