东北大学学报(自然科学版) ›› 2024, Vol. 45 ›› Issue (7): 928-935.DOI: 10.12068/j.issn.1005-3026.2024.07.003
收稿日期:
2023-03-05
出版日期:
2024-07-15
发布日期:
2024-10-29
通讯作者:
李红利
基金资助:
Hong-li LI1(), Hao-yu LIU1, Rong-hua ZHANG2, Yi CHENG1
Received:
2023-03-05
Online:
2024-07-15
Published:
2024-10-29
Contact:
Hong-li LI
About author:
LI Hong-liE-mail:lihongliln@163.com摘要:
情感脑电信号是一种低信噪比的非平稳时间序列,传统的特征提取与分类方法难以提取不同情感状态时的有效特征并进行分类.针对以上情况,提出一种自动融合脑电信号不同频段和时频域特征的深度学习模型.首先,对预处理后的数据进行分频段处理,提取各频段的微分熵特征.然后,在网络中接入的挤压激励模块,对不同频段特征的微分熵特征赋权值,来获取输入数据的有价值信息,再利用改进的稠密连接网络进行特征融合并分类识别,保证了网络层之间最大程度的信息传输.最后,利用SEED情感脑电信号三分类数据集对算法进行了验证,分类正确率可达96.03%,高于现有的基线学习算法,所提算法可进一步增强网络特征提取能力,具有较快的收敛速度,对提升分类器的性能研究具有重要意义.
中图分类号:
李红利, 刘浩雨, 张荣华, 成怡. 基于多维特征矩阵和改进稠密连接网络的情感分类[J]. 东北大学学报(自然科学版), 2024, 45(7): 928-935.
Hong-li LI, Hao-yu LIU, Rong-hua ZHANG, Yi CHENG. Emotional Classification Based on Multidimensional Feature Matrix and Improved Dense Connection Network[J]. Journal of Northeastern University(Natural Science), 2024, 45(7): 928-935.
名称 | 单位样本输入尺寸 | 单位样本输出尺寸 |
---|---|---|
输入 | (8,9,4) | (8,9,4) |
SE模块 | (8,9,4) | (8,9,4) |
卷积层(3×3) | (8,9,4) | (8,9,64) |
稠密连接层1 | (8,9,64) | (8,9,96) |
过渡层1 | (8,9,96) | (4,4,48) |
稠密连接层2 | (4,4,48) | (4,4,80) |
过渡层2 | (4,4,80) | (2,2,40) |
稠密连接层3 | (2,2,40) | (2,2,72) |
SE模块 | (2,2,72) | (2,2,72) |
全局池化层 | (2,2,72) | 72 |
全连接层 | 128 | 3 |
表1 各层网络输入参数尺寸
Table 1 Input parameter size of each layer of network
名称 | 单位样本输入尺寸 | 单位样本输出尺寸 |
---|---|---|
输入 | (8,9,4) | (8,9,4) |
SE模块 | (8,9,4) | (8,9,4) |
卷积层(3×3) | (8,9,4) | (8,9,64) |
稠密连接层1 | (8,9,64) | (8,9,96) |
过渡层1 | (8,9,96) | (4,4,48) |
稠密连接层2 | (4,4,48) | (4,4,80) |
过渡层2 | (4,4,80) | (2,2,40) |
稠密连接层3 | (2,2,40) | (2,2,72) |
SE模块 | (2,2,72) | (2,2,72) |
全局池化层 | (2,2,72) | 72 |
全连接层 | 128 | 3 |
模型 | 正确率 |
---|---|
BiLSTM[ | 90.22 |
DCRN[ | 91.18 |
DAEN[ | 91.01 |
STRNN[ | 89.50 |
SVM, LOSO[ | 83.33 |
MCNN [ | 91.31 |
GELM [ | 91.07 |
DGCNN [ | 90.40 |
本文模型 | 96.03 |
表2 不同模型的分类正确率 (%)
Table 2 Classification accuracy of different models
模型 | 正确率 |
---|---|
BiLSTM[ | 90.22 |
DCRN[ | 91.18 |
DAEN[ | 91.01 |
STRNN[ | 89.50 |
SVM, LOSO[ | 83.33 |
MCNN [ | 91.31 |
GELM [ | 91.07 |
DGCNN [ | 90.40 |
本文模型 | 96.03 |
9通道 | 15通道 | 33通道 | 62通道 | |
---|---|---|---|---|
1 | 74.48 | 87.75 | 92.32 | 93.48 |
2 | 76.09 | 92.14 | 94.75 | 97.33 |
3 | 84.89 | 90.03 | 93.24 | 96.45 |
4 | 86.07 | 97.79 | 99.86 | 99.85 |
5 | 76.67 | 86.54 | 94.57 | 95.41 |
6 | 88.64 | 90.77 | 98.53 | 99.44 |
7 | 80.24 | 86.23 | 93.32 | 92.46 |
8 | 86.45 | 91.64 | 96.54 | 97.33 |
9 | 84.85 | 89.31 | 97.35 | 96.59 |
10 | 86.75 | 91.25 | 97.37 | 97.34 |
11 | 83.43 | 89.42 | 93.36 | 95.56 |
12 | 81.56 | 88.64 | 92.41 | 95.71 |
13 | 83.16 | 89.72 | 96.32 | 95.27 |
14 | 82.48 | 89.93 | 97.78 | 96.44 |
15 | 75.69 | 85.56 | 90.03 | 91.86 |
平均 | 82.10 | 89.78 | 95.18 | 96.03 |
表3 各受试在不同数目电极通道上的分类正确率 (%)
Table 3 Classification accuracy of each subject on different number of electrode channels
9通道 | 15通道 | 33通道 | 62通道 | |
---|---|---|---|---|
1 | 74.48 | 87.75 | 92.32 | 93.48 |
2 | 76.09 | 92.14 | 94.75 | 97.33 |
3 | 84.89 | 90.03 | 93.24 | 96.45 |
4 | 86.07 | 97.79 | 99.86 | 99.85 |
5 | 76.67 | 86.54 | 94.57 | 95.41 |
6 | 88.64 | 90.77 | 98.53 | 99.44 |
7 | 80.24 | 86.23 | 93.32 | 92.46 |
8 | 86.45 | 91.64 | 96.54 | 97.33 |
9 | 84.85 | 89.31 | 97.35 | 96.59 |
10 | 86.75 | 91.25 | 97.37 | 97.34 |
11 | 83.43 | 89.42 | 93.36 | 95.56 |
12 | 81.56 | 88.64 | 92.41 | 95.71 |
13 | 83.16 | 89.72 | 96.32 | 95.27 |
14 | 82.48 | 89.93 | 97.78 | 96.44 |
15 | 75.69 | 85.56 | 90.03 | 91.86 |
平均 | 82.10 | 89.78 | 95.18 | 96.03 |
1 | 郑伟龙,石振锋,吕宝粮.用异质迁移学习构建跨被试脑电情感模型[J].计算机学报,2020,43(2):177-189. |
Zheng Wei‑long, Shi Zhen‑feng, Bao‑liang Lyu.Building cross‑subject EEG‑based affective models using heterogeneous transfer learning[J].Chinese Journal of Computers,2020,43(2):177-189. | |
2 | 权学良,曾志刚,蒋建华,等.基于生理信号的情感计算研究综述[J].自动化学报,2021,47(8):1769-1784. |
Quan Xue‑liang, Zeng Zhi‑gang, Jiang Jian‑hua,et al.Physiological signals based affective computing:a systematic review[J].Acta Automatica Sinica,2021,47(8):1769-1784. | |
3 | Sani O G, Yang Y X, Lee M B,et al.Mood variations decoded from multi‑site intracranial human brain activity[J].Nature Biotechnology,2018,36(10):954-961. |
4 | Zheng W L, Liu W, Lu Y F,et al.EmotionMeter:a multimodal framework for recognizing human emotions[J].IEEE Transactions on Cybernetics,2019,49(3):1110-1122. |
5 | 褚亚奇,朱波,赵新刚,等.基于时空特征学习卷积神经网络的运动想象脑电解码方法[J].生物医学工程学杂志,2021,38(1):1-9. |
Chu Ya‑qi, Zhu Bo, Zhao Xin‑gang,et al.Convolutional neural network based on temporal‑spatial feature learning for motor imagery electroencephalogram signal decoding[J]. Journal of Biomedical Engineering,2021,38(1):1-9. | |
6 | Fdez J, Guttenberg N, Witkowski O,et al.Cross‑subject EEG‑based emotion recognition through neural networks with stratified normalization[J].Frontiers in Neuroscience,2021,15:626277. |
7 | Joshi V M, Ghongade R B.EEG based emotion detection using fourth order spectral moment and deep learning[J].Biomedical Signal Processing and Control,2021,68:102755.1-102755.12. |
8 | 周如双,赵慧琳,林玮玥,等.基于深浅特征融合的深度卷积残差网络的脑电情绪识别模型[J].中国生物医学工程学报,2021,40(6):641-652. |
Zhou Ru‑shuang, Zhao Hui‑lin, Lin Wei‑yue,et al.Feature fusion based deep residual networks using deep and shallow learning for EEG‑based emotion recognition[J].Chinese Journal of Biomedical Engineering,2021,40(6):641-652. | |
9 | Liu W, Zheng W L, Lu B L.Emotion recognition using multimodal deep learning[C]//International Conference on Neural Information Processing.Cham:Springer,2016:521-529. |
10 | Zhang T, Zheng W, Cui Z,et al.Spatial‑temporal recurrent neural network for emotion recognition[J].IEEE Transactions on Cybernetics,2017:839-847. |
11 | Song D W, Zhang Y Z, Hu B,et al.Exploring EEG features in cross‑subject emotion recognition[J].Frontiers in Neuroscience,2018,12:162. |
12 | Duan R N, Zhu J Y, Lu B L.Differential entropy feature for EEG‑based emotion classification[C]//2013 6th International IEEE/EMBS Conference on Neural Engineering (NER).San Diego,2013:1-4. |
13 | 苗敏敏,徐宝国,胡文军,等.基于自适应优化空频微分熵的情感脑电识别[J].仪器仪表学报,2021,42(3):221-230. |
Miao Min‑min, Xu Bao‑guo, Hu Wen‑jun,et al.Emotion EEG recognition based on the adaptive optimized spatial‑frequency differential entropy[J].Chinese Journal of Scientific Instrument,2021,42(3):221-230. | |
14 | Zheng W L, Lu B L.Investigating critical frequency bands and channels for EEG‑based emotion recognition with deep neural networks[J].IEEE Transactions on Autonomous Mental Development,2015,7(3):162-175. |
15 | Shi L C, Jiao Y Y, Lu B L.Differential entropy feature for EEG‑based vigilance estimation[C]//International Conference of the IEEE Engineering in Medicine & Biology Society. San Diego,2013:6627-6630. |
16 | 戴紫玉,马玉良,高云园,等.基于多尺度卷积核CNN的脑电情绪识别[J].传感技术学报,2021,34(4):496-503. |
Dai Zi‑yu, Ma Yu‑liang, Gao Yun‑yuan,et al.A multi‑scale convolutional kernel CNN for EEG emotion recognition[J].Journal of Transduction Technology 2021,34(4):496-503. | |
17 | 李红利,尹飞超,张荣华,等.基于通道注意力和稀疏时频分解的运动想象分类[J].生物医学工程学杂志,2022,39(3):488-497. |
Li Hong‑li, Yin Fei‑chao, Zhang Rong‑hua,et al.Motor imagery electroencephalogram classification based on sparse spatiotemporal decomposition and channel attention[J].Journal of Biomedical Engineering,2022,39(3):488-497. | |
18 | Hu J, Shen L, Sun G.Squeeze‑and‑excitation networks[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Salt Lake City,2018:7132-7141. |
19 | 晁浩,曹益鸣,刘永利.基于三维特征矩阵和冲压激励网络的多通道脑电情感识别[J].控制与决策,2023,38(12):3427-3435. |
Chao Hao, Cao Yi‑ming, Liu Yong‑li.Emotion recognition from multi‑channel EEG data through three‑dimensional feature matrix and squeeze‑and‑excitation networks[J].Control and Decision,2023,38(12):3427-3435. | |
20 | Chen B L, Zhao T S, Liu J H,et al.Multipath feature recalibration DenseNet for image classification[J].International Journal of Machine Learning and Cybernetics,2021,12(3):651-660. |
21 | 柴国强,王大为,芦宾,等.基于注意机制的轻量化稠密连接网络单幅图像去雨[J].北京航空航天大学学报,2022,48(11):2186-2192. |
Chai Guo‑qiang, Wang Da‑wei, Lu Bin,et al.Lightweight dense network based on attention mechanism for single‑image deraining[J].Journal of Beijing University of Aeronautics and Astronautics,2022,48(11):2186-2192. | |
22 | Kingma D P, Ba J.Adam:a method for stochastic optimization[C]// The 3rd International Conference on Learning Representations.San Diego,2015:1412.6980. |
23 | Dang W D, Lyu D M, Li R M,et al.Multilayer network‑based CNN model for emotion recognition[J].International Journal of Bifurcation and Chaos in Applied Sciences and Engineering,2022,32(1):2250011. |
24 | Zheng W L, Zhu J Y, Lu B L.Identifying stable patterns over time for emotion recognition from EEG[J].IEEE Transactions on Affective Computing,2019,10(3):417-429. |
25 | Song T, Zheng W, Song P,et al.EEG emotion recognition using dynamical graph convolutional neural networks[J].IEEE Transactions on Affective Computing,2020,11(3):532-541. |
26 | Zheng W L, Guo H T, Lu B L.Revealing critical channels and frequency bands for emotion recognition from EEG with deep belief network[C]//7th International IEEE/EMBS Conference on Neural Engineering (NER).Montpellier,2015:154-157. |
[1] | 刘纪红, 张律恒, 杨海旭. 一种细胞荧光显微图像饱和伪影修复算法[J]. 东北大学学报(自然科学版), 2024, 45(7): 921-927. |
[2] | 韩东红, 孔彦茹, 展艺萌, 刘源. 音乐多模态数据情感识别方法的研究[J]. 东北大学学报(自然科学版), 2024, 45(6): 776-785. |
[3] | 章伟琪, 王辉明. 混凝土抗压强度的可解释深度学习预测模型[J]. 东北大学学报(自然科学版), 2024, 45(5): 738-744. |
[4] | 尉健一, 吴菁晶. 基于边缘计算的工业物联网中资源分配算法[J]. 东北大学学报(自然科学版), 2023, 44(8): 1072-1078. |
[5] | 杨譞, 何占奇. 改进的两层BiLSTM的心电信号分割方法[J]. 东北大学学报(自然科学版), 2023, 44(12): 1705-1711. |
[6] | 季策 , 王鑫, 耿蓉, 梁敏骏. 时变信道下基于LSTM的信道估计方法[J]. 东北大学学报(自然科学版), 2023, 44(11): 1521-1528. |
[7] | 陈城, 史培新, 王占生, 贾鹏蛟. 基于融合多注意力机制的深度学习的盾构荷载预测方法[J]. 东北大学学报(自然科学版), 2023, 44(11): 1631-1638. |
[8] | 李娟莉, 魏代良, 李博, 文小. 基于深度学习轻量化的改进SSD煤矸快速分选模型[J]. 东北大学学报(自然科学版), 2023, 44(10): 1474-1480. |
[9] | 赵永, 焦诗卉, 赵乾百. 基于Mel频谱和LSTM-DCNN的矿山微震信号混合识别模型[J]. 东北大学学报(自然科学版), 2023, 44(10): 1481-1489. |
[10] | 张雪峰, 王照乙. 基于双决斗深度Q网络的自动换道决策模型[J]. 东北大学学报(自然科学版), 2023, 44(10): 1369-1376. |
[11] | 王瀛, 王泽浩, 李红, 黄文军. 基于深度学习的威胁情报领域命名实体识别[J]. 东北大学学报(自然科学版), 2023, 44(1): 33-39. |
[12] | 顾德英, 罗聿伦, 李文超. 基于改进YOLOv5算法的复杂场景交通目标检测[J]. 东北大学学报(自然科学版), 2022, 43(8): 1073-1079. |
[13] | 代茵, 刘维宾, 董昕阳, 宋雨朦. 基于注意力机制的U-Net脑脊液细胞分割[J]. 东北大学学报(自然科学版), 2022, 43(7): 944-950. |
[14] | 李鸿儒, 任子洋, 黄友鹤, 于霞. 基于变权重奇异谱分析的心律不齐识别方法[J]. 东北大学学报(自然科学版), 2022, 43(3): 305-312. |
[15] | 顾德英, 张松, 孟范伟. 基于单目视觉的车辆3D空间检测方法[J]. 东北大学学报(自然科学版), 2022, 43(3): 328-334. |
阅读次数 | ||||||
全文 |
|
|||||
摘要 |
|
|||||