Two-Level Domain Adaptation Neural Network for EEG-Based Emotion Recognition.
Two-Level Domain Adaptation Neural Network for EEG-Based Emotion Recognition.
复制标题
基于脑电图的情绪识别的两级域适应神经网络
DOI：
10.3389/fnhum.2020.605246
复制
发表时间：
2020
影响因子：
2.9
通讯作者：
Shen Z
中科院分区：
文献类型：
--
作者：
Bao G;Zhuang N;Tong L;Yan B;Shu J;Wang L;Zeng Y;Shen Z
情感识别在人机交互中发挥着重要作用。目前，基于脑电信号（EEG）的情感识别的主要挑战是脑电信号的非平稳性，这导致训练模型的性能随时间而下降。在本文中，我们提出了一个两级域自适应神经网络（TDANN），以构建一个基于EEG的情感识别的传输模型。具体地，使用深度神经网络从拓扑图提取深度特征，其保留来自EEG信号的拓扑信息。然后将这些特征通过TDANN进行两级域混淆。第一层使用最大平均差异（MMD）来减少源域和目标域之间的深度特征分布差异，第二层使用领域对抗神经网络（DANN）来迫使深度特征更接近其对应的类中心。我们在自建数据集和公共数据集SEED上评估了模型的域转移性能。在跨日迁移实验中，准确区分快乐和其他情绪的能力很高：在自建数据集上，悲伤（84%），愤怒（87.04%）和恐惧（85.32%）。在SEED数据集上的准确率达到74.93%。在跨被试迁移实验中，准确区分快乐与其他情绪的能力同样高：在自建数据集上，悲伤（83.79%），愤怒（84.13%）和恐惧（81.72%）。在SEED数据集上的平均准确率达到87.9%，高于WGAN-DA。实验结果表明，该TDANN能有效地处理基于EEG的情感识别中的域迁移问题。
Emotion recognition plays an important part in human-computer interaction (HCI). Currently, the main challenge in electroencephalogram (EEG)-based emotion recognition is the non-stationarity of EEG signals, which causes performance of the trained model decreasing over time. In this paper, we propose a two-level domain adaptation neural network (TDANN) to construct a transfer model for EEG-based emotion recognition. Specifically, deep features from the topological graph, which preserve topological information from EEG signals, are extracted using a deep neural network. These features are then passed through TDANN for two-level domain confusion. The first level uses the maximum mean discrepancy (MMD) to reduce the distribution discrepancy of deep features between source domain and target domain, and the second uses the domain adversarial neural network (DANN) to force the deep features closer to their corresponding class centers. We evaluated the domain-transfer performance of the model on both our self-built data set and the public data set SEED. In the cross-day transfer experiment, the ability to accurately discriminate joy from other emotions was high: sadness (84%), anger (87.04%), and fear (85.32%) on the self-built data set. The accuracy reached 74.93% on the SEED data set. In the cross-subject transfer experiment, the ability to accurately discriminate joy from other emotions was equally high: sadness (83.79%), anger (84.13%), and fear (81.72%) on the self-built data set. The average accuracy reached 87.9% on the SEED data set, which was higher than WGAN-DA. The experimental results demonstrate that the proposed TDANN can effectively handle the domain transfer problem in EEG-based emotion recognition.
登录
查看更多内容
影响因子：
11.3
作者：
GUR, RC;ERWIN, RJ;KRAEMER, HC
通讯作者：
KRAEMER, HC
影响因子：
11.2
作者：
Jenke, Robert;Peer, Angelika;Buss, Martin
通讯作者：
Buss, Martin
影响因子：
2.4
作者：
KOENIG, W;DUNN, HK;LACY, LY
通讯作者：
LACY, LY
基于潜在表示相似性的脑电情感识别领域适应
DOI：
10.1109/tcds.2019.2949306
发表时间：
2020-06-01
影响因子：
5
作者：
Li, Jinpeng;Qiu, Shuang;He, Huiguang
通讯作者：
He, Huiguang
用于脑电情绪识别的双半球域对抗神经网络模型
影响因子：
11.2
作者：
Li, Yang;Zheng, Wenming;Zhou, Xiaoyan
通讯作者：
Zhou, Xiaoyan