A Virtual Reality Art Therapy System Inspired by the Esthetic Conception of Classical Chinese Poetry: Design and Evaluation Based on User Emotions

作者：Tiancheng Jin, Yujie Zhu, Shuai Jia, Jun Wang, Ji Liu · 发表于：International Journal of Human-Computer Interaction · 年份：2025 · DOI：10.1080/10447318.2025.2587928 · 被引用次数：3 · 研究领域：Art Therapy and Mental Health、Advanced Technology in Applications、Aesthetic Perception and Analysis

Emotional distress requires engaging interventions. This study introduces and evaluates an affect-adaptive Virtual Reality (VR) art therapy system leveraging ancient Chinese poetry for emotion regulation. The system integrates AI-generated immersive environments based on a curated poetry-emotion database, uses real-time physiological signals (EDA, HR, IBI) processed via a transfer learning-enhanced model to classify user affect (Valence-Arousal), and dynamically adapts the VR experience accordingly. A pre-post user study (N = 20) revealed significant improvements in self-reported positive emotions (e.g., Joy, Calm) and reductions in negative emotions (e.g., Sadness, Anger; p < 0.05). These findings were corroborated by physiological markers indicating enhanced autonomic regulation (e.g., decreased EDA, increased RMSSD). Users reported high immersion (IPQ) and usability (VRNQ). While realism perception requires refinement and cybersickness mitigation (p < 0.001) is needed, this research demonstrates the feasibility and preliminary effectiveness of using culturally-attuned, affect-adaptive VR for digital art therapy, offering a novel approach to mental well-being support.