模态(人机交互)
人机交互
计算机科学
虚拟现实
多通道交互
增强现实
水准点(测量)
康复
情感计算
触觉技术
可视化
模式治疗法
情绪识别
面部表情
远程康复
多媒体
GSM演进的增强数据速率
可穿戴计算机
虚拟机
用户界面
多模态
人工智能
沉浸式技术
多模式学习
作者
Feng Wang,Jing Nie,Yihan Wang,Jing Xue
出处
期刊:
[Association for Computing Machinery]
日期:2025-10-17
卷期号:: 240-245
标识
DOI:10.1145/3777730.3777770
摘要
Virtual reality (VR) is transitioning from an assistive tool to a core therapeutic modality in rehabilitation and psychological intervention. This paper presents the Immersive Multimodal Adaptive Therapeutic System (IMATS), which integrates multimodal sensing, emotion recognition, and personalized content generation for VR-based art therapy. IMATS employs a CNN–GRU–Attention emotion-recognition model augmented by a quantum support vector machine (QSVM) and couples it with a GAN–Transformer content-generation module to establish a closed-loop “perception–recognition–generation–regulation” architecture. Multimodal inputs comprise electroencephalography (EEG), speech, facial expressions, body movements, heart rate, and haptic feedback. The system is deployed across platforms via the OpenXR standard and achieves low-latency processing through multi-access edge computing (MEC). Benchmark datasets and user studies indicate that IMATS excels in emotion-recognition accuracy, few-shot generalization, immersion, responsiveness, and emotion regulation. The findings suggest that IMATS can serve as an adaptive, user-centered framework for personalized rehabilitative interventions.
科研通智能强力驱动
Strongly Powered by AbleSci AI