[1]杨 青,王亚群,文 斗,等.基于改进 StackCNN 网络和集成学习的脑电信号视觉分类算法[J].郑州大学学报(工学版),2024,45(05):69-76.[doi:10.13705/j.issn.1671-6833.2024.02.009]
　YANG Qing,WANG Yaqun,WEN Dou,et al.EEG Visual Classification Algorithm Based on Improved StackCNN Network andEnsemble Learning[J].Journal of Zhengzhou University (Engineering Science),2024,45(05):69-76.[doi:10.13705/j.issn.1671-6833.2024.02.009]
点击复制

基于改进 StackCNN 网络和集成学习的脑电信号视觉分类算法()
分享到：

《郑州大学学报(工学版)》[ISSN:1671-6833/CN:41-1339/T]

卷:
45
期数:
2024年05期
页码:
69-76
栏目:
出版日期:
2024-08-08

文章信息/Info

Title:
EEG Visual Classification Algorithm Based on Improved StackCNN Network andEnsemble Learning
文章编号:
1671-6833(2024)05-0069-08
作者:
杨 青1; 2; 3;  王亚群1; 2; 3;  文 斗1; 2; 3;  王 莹1; 2; 3;  王翔宇1; 2; 3
1. 华中师范大学 人工智能与智慧学习湖北省重点实验室,湖北 武汉 430079;2. 华中师范大学 计算机学院,湖北武汉 430079;3. 华中师范大学 国家语言资源监测与研究网络媒体中心,湖北 武汉 430079
Author(s):
YANG Qing1; 2; 3;  WANG Yaqun1; 2; 3;  WEN Dou1; 2; 3;  WANG Ying1; 2; 3;  WANG Xiangyu1; 2; 3
1. Hubei Provincial Key Laboratory of Artificial Intelligence and Smart Learning,Central China Normal University, Wuhan 430079,China; 2. School of Computer,Central China Normal University, Wuhan 430079,China; 3. National Language Resources Monitoring &Research Center for Network Media,Central China Normal University, Wuhan 430079,China
关键词:
脑电图;  视觉分类;  卷积神经网络;  Bagging 算法;  ResNet 网络
Keywords:
electroencephagram;  vision classification;  convolutional neural network;  Bagging algorithm;  ResNet network
分类号:
TP399
DOI:
10.13705/j.issn.1671-6833.2024.02.009
文献标志码:
A
摘要:
针对直接使用图像诱发的脑电信号进行视觉分类的现有研究少,并且视觉分类的平均准确率低等问题,设计了一种卷积神经网络( CNN) 和集成学习相结合的方法,用于学习脑电信号相关的视觉特征表示。 通过在StackCNN 网络中加入 K-max 池化方法,解决在提取脑电特征时信息丢失的问题,并结合 Bagging 算法增强网络的泛化能力,该方法称为 StackCNN-B。 采用基于残差神经网络(ResNet)回归对图像进行分类,验证 StackCNN-B 方法在图像分类上的性能。 消融实验及与现有研究对比实验的结果表明:所提方法识别准确率较高,在学习脑电信号的视觉特征表示上的平均准确率达到 99. 78%,在图像分类上的平均准确率达到 96. 45%,与 Bi-LSTM-AttGW 方法相比,平均提高了 0. 28 百分点和 2. 97 百分点。 研究结果验证了脑电信号可以有效地解码与视觉识别相关的人类大脑活动,也表明所提出 StackCNN-B 模型的优越性。
Abstract:
Abstract: Aiming at the limited studies researches on visual classification directly using image-induced EEG signals and low average accuracy of visual classification, a method combining convolutional neural networks ( CNN)and ensemble learning was designed to learn the visual feature representation related to EEG signals. By adding theK-max pooling method to the stackCNN network to solve the problem of information loss when extracting EEGfeatures, and combining with Bagging algorithm to enhance the generalization ability of the network, this methodwas called StackCNN-B. In order to verify the performance of StackCNN-B method in image classification, imageswere classified using deep residual network regression. The results of ablation experiments and comparativeexperiments with existing studies showed that the recognition accuracy of this method was high. The averageaccuracy in learning the visual feature representation of EEG signals was 99. 78%, and the average accuracy inimage classification was 96. 45%. Compared with the most advanced Bi-LSTM-AttGW method, the averageaccuracy was improved by 0. 28 percentage point and 2. 97 percentage point. The results verified that EEG signalscould effectively decode human brain activities related to visual recognition, proved the advantages of the proposedStackCNN-B model.

参考文献/References:

[1] DAS K, GIESBRECHT B, ECKSTEIN M P. Predictingvariations of perceptual performance across individualsfrom neural activity using pattern classifiers[ J] . NeuroImage, 2010, 51(4) : 1425-1437.

[2] CARLSON T, TOVAR D A, ALINK A, et al. Representational dynamics of object vision: the first 1000 ms[ J] .Journal of Vision, 2013, 13(10) : 1.
[3] KANESHIRO B, PERREAU GUIMARAES M, KIM HS, et al. A representational similarity analysis of the dynamics of object processing using single-trial EEG classification[ J] . PLoS One, 2015, 10(8) : e0135697.
[4] EL-LONE R, HASSAN M, KABBARA A, et al. Visualobjects categorization using dense EEG: a preliminarystudy[C]∥2015 International Conference on Advances inBiomedical Engineering ( ICABME) . Piscataway: IEEE,2015: 115-118.
[5] PAREKH V, SUBRAMANIAN R, ROY D, et al. AnEEG-based image annotation system [ M ] ∥Communications in Computer and Information Science. Berlin:Springer, 2018: 303-313.
[6] SPAMPINATO C, PALAZZO S, KAVASIDIS I, et al.Deep learning human mind for automated visual classification[C]∥2017 IEEE Conference on Computer Vision andPattern Recognition (CVPR) . Piscataway: IEEE, 2017:4503-4511.
[7] ZHENG X, CHEN W Z, YOU Y, et al. Ensemble deeplearning for automated visual classification using EEG signals[ J] . Pattern Recognition, 2020, 102: 107147.
[8] ZHENG X, CHEN W Z. An attention-based Bi-LSTM method for visual object classification via EEG[ J]. BiomedicalSignal Processing and Control, 2021, 63: 102174.
[9] MISHRA A, RAJ N, BAJWA G. EEG-based image feature extraction for visual classification using deep learning[C] ∥2022 International Conference on Intelligent DataScience Technologies and Applications ( IDSTA) . Piscataway: IEEE, 2022: 181-188.
[10] KHALEGHI N, REZAII T Y, BEHESHTI S, et al. Developing an efficient functional connectivity-based geometric deep network for automatic EEG-based visual decoding[ J] . Biomedical Signal Processing and Control, 2023,80: 104221.
[11] SHU B, REN F J, BAO Y W. Investigating LSTM withk-max pooling for text classification[C]∥2018 11th International Conference on Intelligent Computation Technology and Automation ( ICICTA ) . Piscataway: IEEE,2018: 31-34.
[12] KIM Y. Convolutional neural networks for sentence classification[ C ] ∥Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing(EMNLP) . Stroudsburg: ACL, 2014: 1746-1751.
[13] GUO B, ZHANG C X, LIU J M, et al. Improving textclassification with weighted word embeddings via a multichannel TextCNN model [ J ] . Neurocomputing, 2019,363: 366-374.
[14] 周澳回, 翁知远, 周思源, 等. 一种基于主题过滤和语义匹配的服务发现方法[ J] . 郑州大学学报( 工学版) , 2022, 43(6) : 36-41, 56.
ZHOU A H, WENG Z Y, ZHOU S Y, et al. A servicediscovery method based on topic filtering and semanticmatching [ J ] . Journal of Zhengzhou University ( Engineering Science) , 2022, 43(6) : 36-41, 56.
[15] MONTANARO A, EBISUZAKI T, BERTAINA M.Stack-CNN algorithm: a new approach for the detection ofspace objects[ J] . Journal of Space Safety Engineering,2022, 9(1) : 72-82.
[16] SIMONYAN K, ZISSERMAN A. Very deep convolutionalnetworks for large-scale image recognition [ EB / OL ] .(2014- 09 - 04) [ 2023 - 07 - 15 ] . http:∥doi. org / 10.48550 / arXiv. 1409. 1556.
[17] BAUER E, KOHAVI R. An empirical comparison of voting classification algorithms: Bagging, boosting, and variants[ J] . Machine Learning, 1999, 36(1) : 105-139.
[18] HE K M, ZHANG X Y, REN S Q, et al. Deep residuallearning for image recognition[C]∥2016 IEEE Conferenceon Computer Vision and Pattern Recognition ( CVPR).Piscataway: IEEE, 2016: 770-778.
[19] WU Z F, SHEN C H, VAN DEN HENGEL A. Wider ordeeper: revisiting the ResNet model for visual recognition[ J] . Pattern Recognition, 2019, 90: 119-133.
[20] DUBEY V K, SARKAR S, SHUKLA R, et al. Epilepticseizure stage classification from EEG signal using ResNet18 model and data augmentation [ C ] ∥2022 IEEEDelhi Section Conference ( DELCON ) . Piscataway:IEEE, 2022: 1-5.
[21] 杨剑锋, 乔佩蕊, 李永梅, 等. 机器学习分类问题及算法研究综述[J]. 统计与决策, 2019, 35(6): 36-40.
YANG J F, QIAO P R, LI Y M, et al. A review of machine-learning classification and algorithms[ J] . Statistics& Decision, 2019, 35(6) : 36-40.
[22] SHAH K, PATEL H, SANGHVI D, et al. A comparativeanalysis of logistic regression, random forest and KNNmodels for the text classification[ J] . Augmented HumanResearch, 2020, 5(1) : 12.
[23] RUSSAKOVSKY O, DENG J, SU H, et al. ImageNetlarge scale visual recognition challenge[ J] . InternationalJournal of Computer Vision, 2015, 115(3) : 211-252.
[24] 郝旺身, 陈耀, 孙浩, 等. 基于全矢-CNN 的轴承故障诊断研 究 [ J] . 郑 州 大 学 学 报 ( 工 学 版) , 2020, 41(5) : 92-96.
HAO W S, CHEN Y, SUN H, et al. Bearing fault diagnosis based on full vector-CNN[J]. Journal of Zhengzhou University (Engineering Science), 2020, 41(5): 92-96.

更新日期/Last Update: 2024-09-02