Immersive Music Therapy Using Virtual Reality and Mobile Technologies
DOI:
https://doi.org/10.3991/ijim.v20i04.60519Keywords:
VR; interactive mobile technologies; immersive music therapy; multimodal emotion recognition; feature-level fusion; model-level fusion; psychological rehabilitation path optimizationAbstract
The global prevalence of mental disorders such as depression and anxiety continues to rise. Traditional music therapy faces challenges such as a lack of immersion, delayed emotional response, and rigid rehabilitation pathways, making it difficult to meet the demand for personalized treatment. The immersive characteristics of virtual reality (VR) and the portability of interactive mobile technologies provide technical support to overcome these limitations, with accurate emotion recognition being a key prerequisite for personalized intervention. This paper aims to construct an immersive music therapy environment based on the deep integration of VR and interactive mobile technologies and proposes a dual-level fusion method for multimodal emotion recognition, combining feature-level and model-level integration to dynamically optimize the psychological rehabilitation path. Methodologically, we first design a “hardware collaboration-software adaptation-interactive feedback loop” architecture that integrates physiological signal acquisition, VR scene rendering, and mobile interaction control modules. Multimodal data, including EEG, ECG, facial expression images, and subjective emotional ratings, are collected. These are then aligned and weighted across modalities at the feature level to extract high-level features, and deep learning models with attention mechanisms at the model level are used for precise emotion classification. Finally, based on real-time emotion recognition results, an optimization algorithm driven by reinforcement learning is developed to dynamically adjust music parameters and VR scene elements. The study confirms that the integration of VR and interactive mobile technologies can break through the limitations of traditional therapy scenarios. The dual-level fusion strategy provides higher accuracy and robustness for emotion recognition, while the dynamic optimization mechanism offers personalized solutions for psychological rehabilitation, with significant academic innovation and clinical application potential.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Sai Wang

This work is licensed under a Creative Commons Attribution 4.0 International License.

