Yong Peng, Fengzhe Jin, Wanzeng Kong et al. (6 total)
2022-05-16
IEEE Transactions on Neural Systems and Rehabilitation Engineering Vol. 30
10.1109/tnsre.2022.3175464
43 citations
摘要
Electroencephalogram(EEG) signals are generated from central nervous system which are difficult to disguise, leading to its popularity in emotion recognition. Recently,semi-supervisedlearning exhibits promisingemotion recognition performance by involving unlabeled EEG data into model training. However, if we first build a graph to characterize the sample similarities and then perform label propagation on this graph, these two steps cannotwell collaborate with each other. In this paper, we propos...
脑电图（EEG）信号在情感识别中越来越受欢迎。半监督学习通过将未标记的脑电图数据纳入模型训练，展现出良好的情感识别性能。然而，图构建和标签传播两个步骤的协同性不足。本研究旨在解决这些问题。
本研究提出了一种最优图耦合半监督学习（OGSSL）模型，用于脑电图情感识别，通过将自适应图学习和情感识别统一到一个目标中。改进了未标记样本的标签指示矩阵，以直接获得它们的情感状态。
在SEED-IV数据集上的实验结果表明，OGSSL在三个跨会话情感识别任务中取得了优异的平均准确率，分别为76.51%、77.08%和81.29%。OGSSL在情感识别中具有区分性的脑电特征选择能力。