论文部分内容阅读
利用功能性近红外光谱(f NIRs)技术实现了对不同情绪状态的识别。采集了15名受试者在6种情绪种类图片刺激下的f NIRs信号以及唤醒度、愉悦度评价数据。为了实现对受试者情绪状态的分类评估,采用支持向量机(SVM)和基于支持向量机的递归特征筛选(SVM-RFE)算法来筛选参数并设计情绪状态的分类器。结果表明在多种情绪种类图片刺激下,受试者出现了显著的功能响应曲线,并且在唤醒度、愉悦度和情绪种类三个分类目标上分别实现了81%、78.78%和68%的平均分类正确率。同时发现唤醒度和愉悦度的敏感特征主要出现在眶额叶皮层和背外侧皮层,且近似熵是反映情绪状态变化的有效指标。因此采用f NIRs能够基本实现对人体情绪状态的识别。
The use of functional near infrared spectroscopy (f NIRs) technology to achieve the recognition of different emotional states. Fifteen subjects were enrolled in the f NIRs signals stimulated by six emotions pictures and the evaluation of arousal and pleasure. In order to classify the emotional state of the subjects, SVM and SVM-RFE algorithms were used to filter the parameters and design the classifier of emotional state. The results showed that under the stimulation of a variety of emotive images, the subjects showed a significant functional response curve and achieved an average of 81%, 78.78% and 68% respectively in the three categories of arousal, pleasure and emotional categories Classification accuracy. At the same time, it is found that the sensitive features of arousal and pleasure mainly appear in the orbitofrontal cortex and dorsolateral cortex, and the approximate entropy is an effective indicator to reflect the change of emotional state. Therefore, the use of f NIRs can basically realize the recognition of the human emotional state.