《计算机应用》唯一官方网站 ›› 2024, Vol. 44 ›› Issue (2): 645-653.DOI: 10.11772/j.issn.1001-9081.2023030286
所属专题: 前沿与综合应用
收稿日期:
2023-03-20
修回日期:
2023-05-06
接受日期:
2023-05-10
发布日期:
2023-05-26
出版日期:
2024-02-10
通讯作者:
罗天健
作者简介:
曹铉(1998—),男,山东兖州人,硕士研究生,主要研究方向:脑机接口、模式识别、脑电信号分析;
基金资助:
Xuan CAO1,2, Tianjian LUO1,2()
Received:
2023-03-20
Revised:
2023-05-06
Accepted:
2023-05-10
Online:
2023-05-26
Published:
2024-02-10
Contact:
Tianjian LUO
About author:
CAO Xuan, born in 1998, M. S. candidate. His research interests include brain computer interface, pattern recognition, EEG signal analysis.
Supported by:
摘要:
解码运动想象脑电(EEG)信号是构造脑机接口(BCI)的关键技术之一。然而,脑电样本采集成本高、个体差异大,且信号具有时变性强、低信噪比等特点,构建跨被试模式识别方法成为了研究的关键。为此,提出一种跨被试动态多域对抗学习方法。首先采用样本协方差对齐和全局域鉴别器适应样本集边缘分布,随后采用多个类别子域鉴别器适应样本集条件分布,并自适应学习多域鉴别器的对抗系数。基于动态多域对抗学习策略,所提出的动态多域对抗网络(DMDAN)模型可学习到被试域间有泛化能力的深度特征。在BCI Competition IV 2A和2B公开数据集上的实验结果表明,DMDAN模型提高了跨被试域不变特征的学习能力,与现有对抗学习方法DRDA(Deep Representation Domain Adaptation)相比,在数据集2A和数据集2B上的平均分类准确率分别提高了1.80和2.52个百分点。可见,所提出的DMDAN模型提升了跨被试运动想象脑电信号解码性能,在不同数据集上具有不错的泛化性。
中图分类号:
曹铉, 罗天健. 运动想象脑电信号的跨被试动态多域对抗学习方法[J]. 计算机应用, 2024, 44(2): 645-653.
Xuan CAO, Tianjian LUO. Dynamic multi-domain adversarial learning method for cross-subject motor imagery EEG signals[J]. Journal of Computer Applications, 2024, 44(2): 645-653.
模块 | 层 | 卷积核 | 卷积核步长 | 卷积核数 |
---|---|---|---|---|
特征 提取器 | 时域特征卷积 | 1×30 | [1,1] | 30 |
空域特征卷积 | C×1 | [1,1] | 30 | |
均值池化 | 1×90 | [ | — | |
分类器 | 全连接层 | 128 | — | — |
全连接层 | 64 | — | — | |
全连接层 | 类别 | — | — |
表1 特征提取器与分类器CNN模型参数
Tab. 1 CNN model parameters of feature extractor and classifier
模块 | 层 | 卷积核 | 卷积核步长 | 卷积核数 |
---|---|---|---|---|
特征 提取器 | 时域特征卷积 | 1×30 | [1,1] | 30 |
空域特征卷积 | C×1 | [1,1] | 30 | |
均值池化 | 1×90 | [ | — | |
分类器 | 全连接层 | 128 | — | — |
全连接层 | 64 | — | — | |
全连接层 | 类别 | — | — |
模块 | 层 | 卷积核 |
---|---|---|
全局域鉴别器 | 全连接层 | 128 |
全连接层 | 64 | |
全连接层 | 32 | |
全连接层 | 1 | |
激活函数sigmoid | — | |
各类别子域鉴别器 | 全连接层 | 128 |
全连接层 | 32 | |
全连接层 | 16 | |
全连接层 | 1 | |
激活函数sigmoid | — |
表2 全局域鉴别器和各类别子域鉴别器CNN模型参数
Tab. 2 CNN model parameters of global discriminator and all local discriminators
模块 | 层 | 卷积核 |
---|---|---|
全局域鉴别器 | 全连接层 | 128 |
全连接层 | 64 | |
全连接层 | 32 | |
全连接层 | 1 | |
激活函数sigmoid | — | |
各类别子域鉴别器 | 全连接层 | 128 |
全连接层 | 32 | |
全连接层 | 16 | |
全连接层 | 1 | |
激活函数sigmoid | — |
数据集 | 被试者数 | MI类别数 | EEG通道数 | 样本时间点数 | 源域训练样本数 | 目标域训练样本数 | 目标域测试样本数 |
---|---|---|---|---|---|---|---|
2A | 9 | 4 | 22 | 1 000 | 4 608 | 288 | 288 |
2B | 9 | 2 | 3 | 1 000 | 5 800 | 400 | 320 |
表3 数据集2A和2B的样本设置情况
Tab. 3 Sample ettings of dataset 2A and 2B
数据集 | 被试者数 | MI类别数 | EEG通道数 | 样本时间点数 | 源域训练样本数 | 目标域训练样本数 | 目标域测试样本数 |
---|---|---|---|---|---|---|---|
2A | 9 | 4 | 22 | 1 000 | 4 608 | 288 | 288 |
2B | 9 | 2 | 3 | 1 000 | 5 800 | 400 | 320 |
数据集 | 被试者 | 训练时间 | 测试时间 |
---|---|---|---|
2A | A01 | 2 301 | 10 |
A02 | 2 369 | 11 | |
A03 | 2 304 | 11 | |
A04 | 2 165 | 10 | |
A05 | 1 861 | 8 | |
A06 | 1 987 | 8 | |
A07 | 1 914 | 7 | |
A08 | 2 441 | 12 | |
A09 | 3 417 | 19 | |
2B | B01 | 1 121 | 7 |
B02 | 1 054 | 30 | |
B03 | 1 057 | 32 | |
B04 | 1 056 | 32 | |
B05 | 1 050 | 7 | |
B06 | 1 077 | 32 | |
B07 | 1 093 | 28 | |
B08 | 1 059 | 5 | |
B09 | 1 138 | 6 |
表4 DMDAN模型在数据集2A、2B上的运行时间 (s)
Tab. 4 Running time of DMDAN model on dataset 2A and 2B
数据集 | 被试者 | 训练时间 | 测试时间 |
---|---|---|---|
2A | A01 | 2 301 | 10 |
A02 | 2 369 | 11 | |
A03 | 2 304 | 11 | |
A04 | 2 165 | 10 | |
A05 | 1 861 | 8 | |
A06 | 1 987 | 8 | |
A07 | 1 914 | 7 | |
A08 | 2 441 | 12 | |
A09 | 3 417 | 19 | |
2B | B01 | 1 121 | 7 |
B02 | 1 054 | 30 | |
B03 | 1 057 | 32 | |
B04 | 1 056 | 32 | |
B05 | 1 050 | 7 | |
B06 | 1 077 | 32 | |
B07 | 1 093 | 28 | |
B08 | 1 059 | 5 | |
B09 | 1 138 | 6 |
模型 | 不同被试者的分类准确率/% | 分类准确率均值±标准差/% | kappa系数 | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
A01 | A02 | A03 | A04 | A05 | A06 | A07 | A08 | A09 | |||
FBCSP[ | 76.00 | 56.50 | 81.25 | 61.00 | 55.00 | 45.25 | 82.75 | 81.25 | 70.75 | 67.75±13.73 | 0.570 0 |
CCSP[ | 84.72 | 52.78 | 80.90 | 59.38 | 54.51 | 49.31 | 88.54 | 71.88 | 56.60 | 66.51±15.13 | 0.553 5 |
SSCSP[ | 76.74 | 58.68 | 81.25 | 57.64 | 38.54 | 48.26 | 76.39 | 79.17 | 78.82 | 66.17±15.75 | 0.548 9 |
SMM[ | 81.94 | 59.38 | 81.60 | 62.85 | 59.03 | 49.36 | 86.11 | 77.78 | 78.47 | 70.72±13.11 | 0.610 0 |
SSMM[ | 82.64 | 60.76 | 85.76 | 67.01 | 58.68 | 54.51 | 90.97 | 81.25 | 79.51 | 73.45±13.32 | 0.646 0 |
C2CM[ | 87.50 | 65.28 | 90.28 | 66.67 | 62.50 | 45.49 | 89.58 | 83.33 | 79.51 | 74.46±15.33 | 0.659 5 |
ConvNet[ | 76.39 | 55.21 | 89.24 | 74.65 | 56.94 | 54.17 | 92.71 | 77.08 | 76.39 | 72.53±14.24 | 0.633 8 |
MI-CNN[ | 73.26 | 28.82 | 89.58 | 68.06 | 26.39 | 28.82 | 75.35 | 78.82 | 77.08 | 60.69±25.17 | 0.475 8 |
DRDA[ | 83.19 | 55.14 | 87.43 | 75.28 | 62.29 | 57.15 | 86.18 | 83.61 | 82.00 | 74.70±12.96 | 0.663 3 |
DJDAN[ | 86.46 | 68.75 | 93.06 | 85.42 | 72.75 | 63.54 | 95.49 | 85.76 | 83.68 | 81.52±10.92 | — |
DMDAN | 84.72 | 61.11 | 92.36 | 69.44 | 62.85 | 57.64 | 88.89 | 85.42 | 86.11 | 76.50±13.57 | 0.686 7 |
表5 DMDAN模型与基线模型在数据集2A上的平均分类准确率和kappa系数对比结果
Tab. 5 Comparison results of average classification accuracy and kappa coefficient between DMDAN model and baseline models on dataset 2A
模型 | 不同被试者的分类准确率/% | 分类准确率均值±标准差/% | kappa系数 | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
A01 | A02 | A03 | A04 | A05 | A06 | A07 | A08 | A09 | |||
FBCSP[ | 76.00 | 56.50 | 81.25 | 61.00 | 55.00 | 45.25 | 82.75 | 81.25 | 70.75 | 67.75±13.73 | 0.570 0 |
CCSP[ | 84.72 | 52.78 | 80.90 | 59.38 | 54.51 | 49.31 | 88.54 | 71.88 | 56.60 | 66.51±15.13 | 0.553 5 |
SSCSP[ | 76.74 | 58.68 | 81.25 | 57.64 | 38.54 | 48.26 | 76.39 | 79.17 | 78.82 | 66.17±15.75 | 0.548 9 |
SMM[ | 81.94 | 59.38 | 81.60 | 62.85 | 59.03 | 49.36 | 86.11 | 77.78 | 78.47 | 70.72±13.11 | 0.610 0 |
SSMM[ | 82.64 | 60.76 | 85.76 | 67.01 | 58.68 | 54.51 | 90.97 | 81.25 | 79.51 | 73.45±13.32 | 0.646 0 |
C2CM[ | 87.50 | 65.28 | 90.28 | 66.67 | 62.50 | 45.49 | 89.58 | 83.33 | 79.51 | 74.46±15.33 | 0.659 5 |
ConvNet[ | 76.39 | 55.21 | 89.24 | 74.65 | 56.94 | 54.17 | 92.71 | 77.08 | 76.39 | 72.53±14.24 | 0.633 8 |
MI-CNN[ | 73.26 | 28.82 | 89.58 | 68.06 | 26.39 | 28.82 | 75.35 | 78.82 | 77.08 | 60.69±25.17 | 0.475 8 |
DRDA[ | 83.19 | 55.14 | 87.43 | 75.28 | 62.29 | 57.15 | 86.18 | 83.61 | 82.00 | 74.70±12.96 | 0.663 3 |
DJDAN[ | 86.46 | 68.75 | 93.06 | 85.42 | 72.75 | 63.54 | 95.49 | 85.76 | 83.68 | 81.52±10.92 | — |
DMDAN | 84.72 | 61.11 | 92.36 | 69.44 | 62.85 | 57.64 | 88.89 | 85.42 | 86.11 | 76.50±13.57 | 0.686 7 |
模型 | 不同被试者的分类准确率/% | 分类准确率均值±标准差/% | kappa系数 | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
B01 | B02 | B03 | B04 | B05 | B06 | B07 | B08 | B09 | |||
FBCSP[ | 70.00 | 60.36 | 60.94 | 97.50 | 93.12 | 80.63 | 78.13 | 92.50 | 86.88 | 80.00±13.85 | 0.600 0 |
CCSP[ | 63.75 | 56.79 | 50.00 | 93.44 | 65.36 | 81.25 | 72.81 | 87.81 | 82.81 | 72.67±14.73 | 0.454 0 |
SSCSP[ | 65.00 | 56.79 | 54.06 | 95.63 | 74.69 | 79.06 | 80.00 | 87.81 | 82.81 | 75.09±13.97 | 0.501 8 |
SMM[ | 67.81 | 51.79 | 53.44 | 93.31 | 82.81 | 74.69 | 72.19 | 82.50 | 75.62 | 72.68±13.55 | 0.454 0 |
SSMM[ | 74.06 | 55.00 | 55.63 | 94.06 | 86.88 | 82.19 | 76.56 | 92.19 | 85.62 | 78.02±14.41 | 0.560 0 |
C2CM[ | 87.50 | 65.28 | 90.28 | 66.67 | 62.50 | 45.49 | 89.58 | 83.33 | 79.51 | 74.46±15.33 | — |
ConvNet[ | 76.56 | 50.00 | 51.56 | 96.88 | 93.13 | 85.31 | 83.75 | 91.56 | 85.62 | 79.37±17.26 | 0.587 5 |
MI-CNN[ | 75.31 | 57.50 | 56.56 | 96.88 | 92.19 | 83.44 | 84.06 | 92.81 | 86.26 | 80.56±14.75 | 0.611 1 |
DRDA[ | 81.37 | 62.86 | 63.63 | 95.94 | 93.56 | 88.19 | 85.00 | 95.25 | 90.00 | 83.98±12.67 | 0.679 6 |
DJDAN[ | 83.44 | 58.57 | 59.06 | 98.13 | 96.56 | 84.38 | 86.52 | 92.81 | 87.81 | 83.03±14.65 | — |
DMDAN | 76.56 | 73.57 | 82.19 | 95.94 | 98.12 | 83.13 | 91.56 | 95.00 | 82.50 | 86.50±8.90 | 0.730 1 |
表6 DMDAN模型与基线模型在数据集2B上的平均分类准确率和kappa系数对比
Tab. 6 Comparison results of average classification accuracy and kappa coefficient between DMDAN model and baseline models on dataset 2B
模型 | 不同被试者的分类准确率/% | 分类准确率均值±标准差/% | kappa系数 | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
B01 | B02 | B03 | B04 | B05 | B06 | B07 | B08 | B09 | |||
FBCSP[ | 70.00 | 60.36 | 60.94 | 97.50 | 93.12 | 80.63 | 78.13 | 92.50 | 86.88 | 80.00±13.85 | 0.600 0 |
CCSP[ | 63.75 | 56.79 | 50.00 | 93.44 | 65.36 | 81.25 | 72.81 | 87.81 | 82.81 | 72.67±14.73 | 0.454 0 |
SSCSP[ | 65.00 | 56.79 | 54.06 | 95.63 | 74.69 | 79.06 | 80.00 | 87.81 | 82.81 | 75.09±13.97 | 0.501 8 |
SMM[ | 67.81 | 51.79 | 53.44 | 93.31 | 82.81 | 74.69 | 72.19 | 82.50 | 75.62 | 72.68±13.55 | 0.454 0 |
SSMM[ | 74.06 | 55.00 | 55.63 | 94.06 | 86.88 | 82.19 | 76.56 | 92.19 | 85.62 | 78.02±14.41 | 0.560 0 |
C2CM[ | 87.50 | 65.28 | 90.28 | 66.67 | 62.50 | 45.49 | 89.58 | 83.33 | 79.51 | 74.46±15.33 | — |
ConvNet[ | 76.56 | 50.00 | 51.56 | 96.88 | 93.13 | 85.31 | 83.75 | 91.56 | 85.62 | 79.37±17.26 | 0.587 5 |
MI-CNN[ | 75.31 | 57.50 | 56.56 | 96.88 | 92.19 | 83.44 | 84.06 | 92.81 | 86.26 | 80.56±14.75 | 0.611 1 |
DRDA[ | 81.37 | 62.86 | 63.63 | 95.94 | 93.56 | 88.19 | 85.00 | 95.25 | 90.00 | 83.98±12.67 | 0.679 6 |
DJDAN[ | 83.44 | 58.57 | 59.06 | 98.13 | 96.56 | 84.38 | 86.52 | 92.81 | 87.81 | 83.03±14.65 | — |
DMDAN | 76.56 | 73.57 | 82.19 | 95.94 | 98.12 | 83.13 | 91.56 | 95.00 | 82.50 | 86.50±8.90 | 0.730 1 |
模型 | 数据集2A | 数据集2B |
---|---|---|
实验1 | 74.48 | 84.77 |
实验2 | 75.36 | 85.52 |
实验3 | 75.46 | 84.88 |
DMDAN | 76.50 | 86.50 |
表7 不同消融实验设置在两个数据集上平均分类准确率 (%)
Tab. 7 Average classification accuracy of different ablation experiment settings on two datasets
模型 | 数据集2A | 数据集2B |
---|---|---|
实验1 | 74.48 | 84.77 |
实验2 | 75.36 | 85.52 |
实验3 | 75.46 | 84.88 |
DMDAN | 76.50 | 86.50 |
1 | AL-SAEGH A, DAWWD S A, ABDUL-JABBAR J M. Deep learning for motor imagery EEG-based classification: a review [J]. Biomedical Signal Processing and Control, 2021, 63: 102172. 10.1016/j.bspc.2020.102172 |
2 | HOUSSEIN E H, HAMMAD A, ALI A A. Human emotion recognition from EEG-based brain-computer interface using machine learning: a comprehensive review [J]. Neural Computing and Applications, 2022, 34: 12527-12557. 10.1007/s00521-022-07292-4 |
3 | 刘拓, 叶阳阳, 王坤, 等.运动想象脑电信号分类算法的研究进展[J].生物医学工程学杂志, 2021, 38(5): 995-1002. 10.7507/1001-5515.202101089 |
LIU T, YE Y Y, WANG K, et al. Progress of classification algorithms for motor imagery electroencephalogram signals [J]. Journal of Biomedical Engineering, 2021, 38(5): 995-1002. 10.7507/1001-5515.202101089 | |
4 | ARPAIA P, ESPOSITO A, NATALIZIO A, et al. How to successfully classify EEG in motor imagery BCI: a metrological analysis of the state of the art [J]. Journal of Neural Engineering, 2022, 19: 031002. 10.1088/1741-2552/ac74e0 |
5 | 彭禹, 宋耀莲, 杨俊. 基于数据增强的运动想象脑电分类[J].计算机应用, 2022, 42(11): 3625-3632. 10.11772/j.issn.1001-9081.2021091701 |
PENG Y, SONG Y L, YANG J. Motor imagery electroencephalography classification based on data augmentation [J]. Journal of Computer Applications, 2022, 42(11): 3625-3632. 10.11772/j.issn.1001-9081.2021091701 | |
6 | 霍首君, 郝琰, 石慧宇, 等.基于深度卷积网络的运动想象脑电信号模式识别[J]. 计算机应用, 2021, 41(4): 1042-1048. 10.11772/j.issn.1001-9081.2020081300 |
HUO S J, HAO Y, SHI H Y, et al. Pattern recognition of motor imagery EEG based on deep convolutional network [J], Journal of Computer Applications, 2021, 41(4): 1042-1048. 10.11772/j.issn.1001-9081.2020081300 | |
7 | 张宪法, 郝矿荣, 陈磊.免疫多域特征融合的多核学习SVM运动想象脑电信号分类[J]. 自动化学报, 2020, 46(11): 2417-2426. 10.16383/j.aas.c180247 |
ZHANG X F, HAO K R, CHEN L. Motor imagery EEG classification based on immune multi-domain-feature fusion and multiple kernel learning SVM [J]. Acta Automatica Sinica, 2020,46(11): 2417-2426. 10.16383/j.aas.c180247 | |
8 | ANG K K, CHIN Z Y, WANG C, et al. Filter bank common spatial pattern algorithm on BCI competition IV datasets 2a and 2b[J]. Frontiers in Neuroscience, 2012, 6: No.39. 10.3389/fnins.2012.00039 |
9 | KANG H, NAM Y, CHOI S. Composite common spatial pattern for subject-to-subject transfer [J]. IEEE Signal Processing Letters, 2009, 16(8): 683-686. 10.1109/lsp.2009.2022557 |
10 | SAMEK W, MEINECKE F C, K-R MÜLLER. Transferring subspaces between subjects in brain-computer interfacing [J]. IEEE Transactions on Biomedical Engineering, 2013, 60(8): 2289-2298. 10.1109/tbme.2013.2253608 |
11 | 罗天健, 周昌乐.重叠特征策略与参数优化的运动想象脑电模式识别[J].模式识别与人工智能, 2020, 33(8): 692-704. 10.16451/j.cnki.issn1003-6059.202008003 |
LUO T J, ZHOU C L. Overlapped features strategy and parameters optimization patterns recognition for motor imagery EEG [J]. Pattern Recognition and Artificial Intelligence, 2020, 33(8): 692-704. 10.16451/j.cnki.issn1003-6059.202008003 | |
12 | ZHENG Q, ZHU F, QIN J, et al. Sparse support matrix machine [J]. Pattern Recognition, 2018, 76: 715-726. 10.1016/j.patcog.2017.10.003 |
13 | SCHIRRMEISTER R T, SPRINGENBERG T J, FIEDERER L D J, et al. Deep learning with convolutional neural networks for EEG decoding and visualization[J]. Human Brain Mapping, 2017, 38(11): 5391-5420. 10.1002/hbm.23730 |
14 | SAKHAVI S, GUAN C, YAN S. Learning temporal information for brain-computer interface using convolutional neural networks [J]. IEEE Transactions on Neural Networks and Learning Systems, 2018, 29(11): 5619-5629. 10.1109/tnnls.2018.2789927 |
15 | DOSE H, MØLLER J S, IVERSEN H K, et al. An end-to-end deep learning approach to MI-EEG signal classification for BCIs [J]. Expert Systems with Applications, 2018. 114: 532-542. 10.1016/j.eswa.2018.08.031 |
16 | 韦泓妤, 陈黎飞, 罗天健.运动想象脑电信号的跨域特征学习方法[J].计算机应用研究, 2022, 39(8): 2340-2346. 10.11772/j.issn.1001-9081.2023030286 |
WEI H Y, CHEN L F, LUO T J. Cross-domain feature learning method for motor imagery EEG signals [J]. Application Research of Computers, 2022, 39(8): 2340-2346. 10.11772/j.issn.1001-9081.2023030286 | |
17 | ZANINI P, CONGEDO M, JUTTEN C, et al. Transfer learning: A Riemannian geometry framework with applications to brain-computer interfaces [J]. IEEE Transactions on Biomedical Engineering, 2017, 65(5): 1107-1116. 10.1109/tbme.2017.2742541 |
18 | HE H, WU D. Transfer learning for brain-computer interfaces: a Euclidean space data alignment approach [J]. IEEE Transactions on Biomedical Engineering, 2020, 67(2): 399-410. 10.1109/tbme.2019.2913914 |
19 | ZHANG X, SHE Q, CHEN Y, et al. Sub-band target alignment common spatial pattern in brain-computer interface [J]. Computer Methods and Programs in Biomedicine, 2021, 207: 106150. 10.1016/j.cmpb.2021.106150 |
20 | ZHANG W, WU D. Manifold embedded knowledge transfer for brain-computer interfaces [J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2020, 28(5): 1117-1127. 10.1109/tnsre.2020.2985996 |
21 | CAI Y, SHE Q, JI J, et al. Motor imagery EEG decoding using manifold embedded transfer learning [J]. Journal of Neuroscience Methods, 2022, 370: 109489. 10.1016/j.jneumeth.2022.109489 |
22 | GANIN Y, LEMPITSKY V. Unsupervised domain adaptation by back propagation [C]// Proceedings of the 32nd International Conference on International Conference on Machine Learning. New York: JMLR.org, 2015: 1180-1189. |
23 | SAKHAVI S, GUAN C. Convolutional neural network-based transfer learning and knowledge distillation using multi-subject data in motor imagery BCI [C] // Proceedings of the 2017 8th International IEEE/EMBS Conference on Neural Engineering. Piscataway: IEEE, 2017: 588-591. 10.1109/ner.2017.8008420 |
24 | LI Y, ZHENG W, ZONG Y, et al. A bi-hemisphere domain adversarial neural network model for EEG emotion recognition [J]. IEEE Transactions on Affective Computing, 2021, 12(2): 494-504. 10.1109/taffc.2018.2885474 |
25 | ZHAO H, ZHENG Q, MA K, et al. Deep representation-based domain adaptation for nonstationary EEG classification [J]. IEEE Transactions on Neural Networks and Learning Systems, 2021, 32(2): 535-545. 10.1109/tnnls.2020.3010780 |
26 | HONG X, ZHENG Q, LIU L, et al. Dynamic joint domain adaptation network for motor imagery classification [J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2021, 29: 556-565. 10.1109/tnsre.2021.3059166 |
27 | 张玉红,余道远,胡学钢.领域自适应任务中的动态参数调整方法[J].模式识别与人工智能,2021, 34(10): 924-931. 10.16451/j.cnki.issn1003-6059.202110005 |
ZHANG Y H, YU D Y, HU X G. Dynamic parameter setting method for domain adaptation [J]. Pattern Recognition and Artificial Intelligence, 2021, 34(10): 924-931. 10.16451/j.cnki.issn1003-6059.202110005 | |
28 | YU C, WANG J, CHEN Y, et al. Transfer Learning with Dynamic Adversarial Adaptation Network [C]// Proceedings of 2019 IEEE International Conference on Data Mining. Piscataway: IEEE, 2019: 778-786. 10.1109/icdm.2019.00088 |
29 | TANGERMANN M, MÜLLER K R, AERTSEN A, et al. Review of the BCI competition IV [J]. Frontiers in NeuroScience, 2012, 6: No.55. 10.3389/fnins.2012.00055 |
30 | VAN DER MAATEN L, HINTON G. Visualizing data using t-SNE [J]. Journal of Machine Learning Research, 2008, 9: 2579-2605. |
[1] | 晁浩, 封舒琪, 刘永利. 脑电情感识别中多上下文向量优化的卷积递归神经网络[J]. 《计算机应用》唯一官方网站, 2024, 44(7): 2041-2046. |
[2] | 时旺军, 王晶, 宁晓军, 林友芳. 小样本场景下的元迁移学习睡眠分期模型[J]. 《计算机应用》唯一官方网站, 2024, 44(5): 1445-1451. |
[3] | 杨思琪, 罗天健, 严宣辉, 杨光局. 运动想象脑电图的空域特征迁移核学习方法[J]. 《计算机应用》唯一官方网站, 2024, 44(11): 3354-3363. |
[4] | 杨淑莹, 国海铭, 李欣. 基于通道选择和多维特征融合的脑电信号分类[J]. 《计算机应用》唯一官方网站, 2023, 43(11): 3418-3427. |
[5] | 张军鹏, 施玉杰, 蒋睿, 董静静, 邱昌建. 基于脑电信号的认知功能障碍识别与分类进展综述[J]. 《计算机应用》唯一官方网站, 2023, 43(10): 3297-3308. |
[6] | 陈代丽, 许国良. 基于注意力机制学习域内变化的跨域行人重识别方法[J]. 《计算机应用》唯一官方网站, 2022, 42(5): 1391-1397. |
[7] | 秦静, 孙法莉, HUI Fang, 汪祖民, 高兵, 季长清. 可穿戴脑电图设备关键技术及其应用综述[J]. 《计算机应用》唯一官方网站, 2022, 42(4): 1029-1035. |
[8] | 彭禹, 宋耀莲, 杨俊. 基于数据增强的运动想象脑电分类[J]. 《计算机应用》唯一官方网站, 2022, 42(11): 3625-3632. |
[9] | 霍首君, 郝琰, 石慧宇, 董艳清, 曹锐. 基于深度卷积网络的运动想象脑电信号模式识别[J]. 计算机应用, 2021, 41(4): 1042-1048. |
[10] | 袁园, 吴文, 万毅. 基于熵驱动域适应学习的单幅图像阴影检测方法[J]. 计算机应用, 2020, 40(7): 2131-2136. |
[11] | 肖禾, 刘志勤, 王庆凤, 黄俊, 周莹, 刘启榆, 徐卫云. 基于迁移学习的多视角乳腺肿块和钙化簇分类方法[J]. 计算机应用, 2020, 40(5): 1460-1464. |
[12] | 罗飞, 刘鹏飞, 罗元, 朱思蒙. 多特征融合的运动想象脑电特征提取方法[J]. 《计算机应用》唯一官方网站, 2020, 40(2): 616-620. |
[13] | 胡章芳, 张力, 黄丽嘉, 罗元. 基于时频域的卷积神经网络运动想象脑电信号识别方法[J]. 计算机应用, 2019, 39(8): 2480-2483. |
[14] | 谭瑶, 饶文碧. 异构复合迁移学习的视频内容标注方法[J]. 计算机应用, 2018, 38(6): 1547-1553. |
[15] | 胡敏, 李冲, 路荣荣, 黄宏程. 基于三维视觉指导的运动想象训练性能分析[J]. 计算机应用, 2018, 38(3): 836-841. |
阅读次数 | ||||||
全文 |
|
|||||
摘要 |
|
|||||