随着计算技术和人机交互技术的发展，情绪识别成为情绪研究的新兴领域，而人体生理信号与人的情绪存在某种内在的联系，且相对客观、不易伪装，基于生理信号的情绪识别成为近年来情绪识别研究的热点。传统情绪诱导材料存在生态效度低，真实感不强和交互性差等问题，同时，多源生理数据的融合带来信息冗余和噪声大，成为情绪识别研究中的亟待解决的技术难点。本项目拟借助虚拟现实（VR）技术，利用其沉浸式体验，尝试建立虚拟场景与情绪的对应关系，快速准确地诱发受试者相关的情绪，以获得相对客观、低噪声的脑电图（EEG），额肌电（EMG）、心率（HR）和体温（SKT）等多生理数据。同时，提出多核粗糙模糊聚类方法和基于深度置信网络（Deep Belief Network）模型，解决多源数据的异构问题和信息融合中冗余和噪声大等问题，有效提高基于VR场景和多源生理信号交互的情绪识别效果。本研究具有明确的研究思路和广泛的应用。

Emotion recognition is a new field of emotion research along with the development of computing technology and human machine interaction technology. As physiological signals are correlated with human emotion and can hardly be masked，recognizing emotions based on physiological signals has become a research hotspot in recent years．Lacking ecological validity, poor sense of reality, and less interactivity are existing problems to induce emotions in traditional methods. Also, information fusion based on multi-channel physiological signals will bring redundant information and big noise, which have become nasty solvable problems. This project is intended to apply Virtual Reality (VR) technology to design various Virtual scenes. Taking the advantage of immersive experience, we expect to set up a corresponding relationship between VR senses and emotions. Then, we can induce related emotion precisely and rapidly, in order to obtain relatively objective multi-channel physiological signals, including electroencephalogram (EEG), electromyographic signal (EMG), heart rate (HR), and skin-temperature (SKT), with lower noisy. Meanwhile, based on Deep Belief Network and multiple kernel rough fuzzy clustering theory, provide an innovative solution to the problem of heterogeneity, redundancy and noise problem in the process of muti-feature fusion. During this process, a new emotion recognition and VR scene interaction model will be developed. This project has clear ideas and extensive application prospect. It can also provide a new thought in building VR emotional scene database in future.