Comparing user-dependent and user-independent training of CNN for SSVEP BCI
Open accessThe DOI matches its Crossref record; 47 of the 47 references it cites were found in a registry or an index.
Abstract
OBJECTIVE: We presented a comparative study on the training methodologies of a convolutional neural network (CNN) for the detection of steady-state visually evoked potentials (SSVEP). Two training scenarios were also compared: user-independent (UI) training and user-dependent (UD) training. APPROACH: The CNN was trained in both UD and UI scenarios on two types of features for SSVEP classification: magnitude spectrum features (M-CNN) and complex spectrum features (C-CNN). The canonical correlation analysis (CCA), widely used in SSVEP processing, was used as the baseline. Additional comparisons were performed with task-related components analysis (TRCA) and filter-bank canonical correlation analysis (FBCCA). The performance of the proposed CNN pipelines, CCA, FBCCA and TRCA were evaluated with two datasets: a seven-class SSVEP dataset collected on 21 healthy participants and a twelve-class publicly available SSVEP dataset collected on ten healthy participants. MAIN RESULTS: The UD based training methods consistently outperformed the UI methods when all other conditions were the same, as one would expect. However, the proposed UI-C-CNN approach performed similarly to the UD-M-CNN across all cases investigated on both datasets. On Dataset 1, the average accuracies of the different methods for 1 s window length were: CCA: 69.1% ± 10.8%, TRCA: 13.4% ± 1.5%, FBCCA: 64.8% ± 15.6%, UI-M-CNN: 73.5% ± 16.1%, UI-C-CNN: 81.6% ± 12.3%, UD-M-CNN: 87.8% ± 7.6% and UD-C-CNN: 92.5% ± 5%. On Dataset 2, the average accuracies of the different methods for data length of 1 s were: UD-C-CNN: 92.33% ± 11.1%, UD-M-CNN: 82.77% ± 16.7%, UI-C-CNN: 81.6% ± 18%, UI-M-CNN: 70.5% ± 22%, FBCCA: 67.1% ± 21%, CCA: 62.7% ± 21.5%, TRCA: 40.4% ± 14%. Using t-SNE, visualizing the features extracted by the CNN pipelines further revealed that the C-CNN method likely learned both the amplitude and phase related information from the SSVEP data for classification, resulting in superior performance than the M-CNN methods. The results suggested that UI-C-CNN method proposed in this study offers a good balance between performance and cost of training data. SIGNIFICANCE: The proposed C-CNN based method is a suitable candidate for SSVEP-based BCIs and provides an improved performance in both UD and UI training scenarios.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Figures







Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
(2008) · Journal of Machine Learning Research · Cited by 33,051
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
Ioffe, S., & Szegedy, C. (2015) · arXiv · doi:10.48550/arxiv.1502.03167 · Cited by 27,818
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
Ioffe, S. (2024) · arXiv (Cornell University) · doi:10.57702/o9raffed · Cited by 14,214
EEGNet: a compact convolutional neural network for EEG-based brain-computer interfaces
Lawhern, V. J., Solon, A. J., Waytowich, N. R., Gordon, S. M., Hung, C. P., & Lance, B. J. (2018) · Journal of Neural Engineering · doi:10.1088/1741-2552/aace8c · Cited by 5,115 · cited 4×
“%) achieves the highest performance compared to CCA (62.7% ± 21.5%), UI-M-CNN (70.5% ± 22%), FBCCA (67.1% ± 21%), Compact-CNN (79% ± 15”
MethodInfluentialEEGNet: A Compact Convolutional Network for EEG-based Brain-Computer Interfaces
Lawhern, V. J., Solon, A. J., Waytowich, N. R., Gordon, S. M., Hung, C. P., & Lance, B. J. (2016) · arXiv · doi:10.48550/arxiv.1611.08024 · Cited by 5,105 · cited 4×
“%) achieves the highest performance compared to CCA (62.7% ± 21.5%), UI-M-CNN (70.5% ± 22%), FBCCA (67.1% ± 21%), Compact-CNN (79% ± 15”
MethodInfluentialA review of classification algorithms for EEG-based brain–computer interfaces: A 10 year update
Lotte, F., Bougrain, L., Cichocki, A., Clerc, M., Congedo, M., Rakotomamonjy, A., & Yger, F. (2018) · Journal of Neural Engineering · doi:10.1088/1741-2552/aab2f2 · Cited by 2,239 · cited 1×
“In a recent study, it was reported that a vast majority of published studies based on deep learning for EEG based BCIs did not compare the proposed techniques to state-of-the-art methods or they performed biased comparisons [40].”
BackgroundDeep learning for electroencephalogram (EEG) classification tasks: a review
Craik, A., He, Y., & Contreras-Vidal, J. L. (2019) · Journal of Neural Engineering · doi:10.1088/1741-2552/ab0ab5 · Cited by 1,758
Deep learning-based electroencephalography analysis: a systematic review
Roy, Y., Banville, H., Albuquerque, I., Gramfort, A., Falk, T. H., & Faubert, J. (2019) · Journal of Neural Engineering · doi:10.1088/1741-2552/ab260c · Cited by 1,532 · cited 1×
“One of the key challenges highlighted in deep learning based methods for BCIs is the reproducibility of results [18], where the authors provided guidelines such as: providing a clear description of the architecture, clearly describing the data used, use of existing datasets and evaluating the performance with baseline.”
BackgroundDeep learning for healthcare applications based on physiological signals: A review
Faust, O., Hagiwara, Y., Hong, T. J., Lih, O. S., & Acharya, U. R. (2018) · Computer Methods and Programs in Biomedicine · doi:10.1016/j.cmpb.2018.04.005 · Cited by 1,026
High-speed spelling with a noninvasive brain–computer interface
Chen, X., Wang, Y., Nakanishi, M., Gao, X., Jung, T.-P., & Gao, S. (2015) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1508080112 · Cited by 969
Cited by
Summary of over Fifty Years with Brain-Computer Interfaces—A Review
Kawala-Sterniuk, A., Browarska, N., Al-Bakri, A., Pelc, M., Zygarlicki, J., Sidikova, M., Martinek, R., & Gorzelanczyk, E. J. (2021) · Brain Sciences · doi:10.3390/brainsci11010043 · Cited by 292
“The SSVEP-BCIs also require little or not training, what makes it a great candidate for real-life applications [113,118].”
BackgroundA transformer-based deep neural network model for SSVEP classification
Chen, J., Zhang, Y., Pan, Y., Xu, P., & Guan, C. (2023) · Neural Networks · doi:10.1016/j.neunet.2023.04.045 · Cited by 170
A Deep Neural Network for SSVEP-Based Brain-Computer Interfaces
Güney, O. B., Oblokulov, M., & Özkan, H. (2021) · IEEE Transactions on Biomedical Engineering · doi:10.1109/tbme.2021.3110440 · Cited by 153
“Indeed, subject specific adaptation has been observed in [39] to largely improve the identification performance.”
BackgroundKohli, V., Tripathi, U., Chamola, V., Rout, B. K., & Kanhere, S. S. (2021) · Microprocessors and Microsystems · doi:10.1016/j.micpro.2021.104392 · Cited by 151
Review of brain encoding and decoding mechanisms for EEG-based brain–computer interface
Xu, L., Xu, M., Jung, T., & Ming, D. (2021) · Cognitive Neurodynamics · doi:10.1007/s11571-021-09676-z · Cited by 112
Pan, Y., Chen, J., Zhang, Y., & Zhang, Y. (2022) · Journal of Neural Engineering · doi:10.1088/1741-2552/ac8dc5 · Cited by 92
Ding, W., Shan, J., Fang, B., Wang, C., Sun, F., & Li, X. (2021) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2021.3132162 · Cited by 78
An Analysis of Deep Learning Models in SSVEP-Based BCI: A Survey
Xu, D., Tang, F., Li, Y., Zhang, Q., & Feng, X. (2023) · Brain Sciences · doi:10.3390/brainsci13030483 · Cited by 68
“Ravi demonstrated that UD-based training methods consistently outperformed UI methods when all other conditions were the same [51].”
MethodLee, D.-Y., Jeong, J.-H., Lee, B.-H., & Lee, S. (2022) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2022.3143836 · Cited by 62
Similarity-constrained task-related component analysis for enhancing SSVEP detection
Sun, Q., Chen, M., Zhang, L., Li, C., & Kang, W. (2021) · Journal of Neural Engineering · doi:10.1088/1741-2552/abfdfa · Cited by 54
Kwon, J., & Im, C. (2021) · Frontiers in Human Neuroscience · doi:10.3389/fnhum.2021.646915 · Cited by 55
“The normalization and dropout layers were added after the input layer and the two convolutional layers to improve the generalization performance and training speed of the networks (Ravi et al., 2020).”
MethodA portable SSVEP-BCI system for rehabilitation exoskeleton in augmented reality environment
Wang, F., Wen, Y., Bi, J., Li, H., & Sun, J. (2023) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2023.104664 · Cited by 51
Automated detection of driver fatigue from electroencephalography through wavelet-based connectivity
Ahmadi, A., Bazregarzadeh, H., & Kazemi, K. (2020) · Journal of Applied Biomedicine · doi:10.1016/j.bbe.2020.08.009 · Cited by 46
TRCA-Net: using TRCA filters to boost the SSVEP classification with convolutional neural network
Deng, Y., Sun, Q., Wang, C., Wang, Y., & Zhou, S. K. (2023) · Journal of Neural Engineering · doi:10.1088/1741-2552/ace380 · Cited by 53
VR-enabled portable brain-computer interfaces via wireless soft bioelectronics
Mahmood, M., Kim, N., Mahmood, M., Kim, H., Kim, H., Rodeheaver, N., Sang, M., Yu, K. J., & Yeo, W. (2022) · Biosensors and Bioelectronics · doi:10.1016/j.bios.2022.114333 · Cited by 47
Apicella, A., Arpaïa, P., Benedetto, E. D., Donato, N., Duraccio, L., Giugliano, S., & Prevete, R. (2022) · IEEE Sensors Journal · doi:10.1109/jsen.2022.3161743 · Cited by 44
Ravi, A., Lu, J., Pearce, S., & Jiang, N. (2022) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2022.3140772 · Cited by 43
Self-citationA hybrid BCI combining SSVEP and EOG and its application for continuous wheelchair control
Mai, X., Ai, J., Ji, M., Zhu, X., & Meng, J. (2023) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2023.105530 · Cited by 44
FB-EEGNet: A fusion neural network across multi-stimulus for SSVEP target detection
Yao, H., Liu, K., Deng, X., Tang, X., & Yu, H. Z. (2022) · Journal of Neuroscience Methods · doi:10.1016/j.jneumeth.2022.109674 · Cited by 43
Filter Bank Convolutional Neural Network for SSVEP Classification
Zhao, D., Wang, T., Tian, Y., & Jiang, X. (2021) · IEEE Access · doi:10.1109/access.2021.3124238 · Cited by 35
“stable during stimulation, Fast Fourier Transform (FFT) [9], [21], [22], [23] is widely applied to transform SSVEP signals into the frequency domain before feeding into classifiers.”
MethodInfluentialIENet: a robust convolutional neural network for EEG based brain-computer interfaces
Du, Y., & Liu, J. (2022) · Journal of Neural Engineering · doi:10.1088/1741-2552/ac7257 · Cited by 34
“Ravi et al proposed taking complex fast Fourier transform (FFT) as input to one-dimensional convolutions for SSVEP classification [39].”
BackgroundMethodInfluentialA Deep Learning Approach for Brain Computer Interaction-Motor Execution EEG Signal Classification
ElSayed, N. E., Tolba, A. S., Rashad, M. Z., Belal, T. M. I., & Sarhan, S. S. (2021) · IEEE Access · doi:10.1109/access.2021.3097797 · Cited by 38
“Finally, UI involves training data from several individuals, after which a generalized model, called a ‘‘UI’’, developed to be applied to unseen users [39].”
BackgroundTransfer learning of an ensemble of DNNs for SSVEP BCI spellers without user-specific training
Güney, O. B., & Özkan, H. (2023) · Journal of Neural Engineering · doi:10.1088/1741-2552/acacca · Cited by 41
Kwon, J., & Im, C. (2022) · Expert Systems with Applications · doi:10.1016/j.eswa.2022.117574 · Cited by 32
Wang, Z., Wong, C. M., Wang, B., Feng, Z., Cong, F., & Wan, F. (2023) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2023.3276745 · Cited by 30
“According to the DNN-based SSVEP recognition methods [6], [22], [23], the recognition performance may be further improved by using calibration”
BackgroundMethodHong, J., & Qin, X. (2021) · Journal of Intelligent & Fuzzy Systems · doi:10.3233/jifs-201280 · Cited by 35
“This method extracts task related components by maximizing their reproducibility during task periods [6, 75].”
Ke, Y., Du, J., Liu, S., & Ming, D. (2023) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2023.3246359 · Cited by 27
“deep learning-based [9], [10], [11], [12] target identification algorithms.”
Representative-Based Cold Start for Adaptive SSVEP-BCI
Shi, N., Li, X., Liu, B., Yang, C., Wang, Y., & Gao, X. (2023) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2023.3245654 · Cited by 28
“Hence, these algorithms can be referred as user-dependent (UD) [4].”
BackgroundDeep transfer learning-based SSVEP frequency domain decoding method
Xiong, H., Song, J., Liu, J., & Han, Y. (2024) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2023.105931 · Cited by 25
Convolutional Transformer-Based Cross Subject Model for SSVEP-Based BCI Classification
Liu, J., Wang, R., Yang, Y., Zong, Y., Leng, Y., Zheng, W., & Ge, S. (2024) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2024.3454158 · Cited by 24
“Although the calibration-free solution for new users is the most convenient in practical applications, the existence of individual differences between subjects means that significant improvements in SSVEP classification after subject-specific calibration are often obtained [18], [48].”
BackgroundBrain–Computer Interface—A Brain-in-the-Loop Communication System
Gao, X., Wang, Y., Chen, X., Liu, B., & Gao, S. (2025) · Proceedings of the IEEE · doi:10.1109/jproc.2025.3600389 · Cited by 32
“…subspace enables various EEG data processing operations, including smoothing, averaging, extrapolating, interpolating, and classifying [181], [207]. d) Deep learning: Deep learning learns the features and classifiers simultaneously in an end-to-end manner [208], [209], [210], [211], [212], [213].”
IoT-driven augmented reality and virtual reality systems in neurological sciences
Sahu, M., Gupta, R. V., Ambasta, R. K., & Kumar, P. (2024) · Internet of Things · doi:10.1016/j.iot.2024.101098 · Cited by 27
Short-length SSVEP data extension by a novel generative adversarial networks based framework
Pan, Y., Li, N., Zhang, Y., Xu, P., & Yao, D. (2024) · Cognitive Neurodynamics · doi:10.1007/s11571-024-10134-9 · Cited by 25
“…in the short-length SSVEPs could be moderately alleviated via studying the subject-invariant or subject-variant properties (Bin et al. 2011; Zhang et al. 2011; Nakanishi et al. 2017; Zhang et al. 2018; Waytowich et al. 2018; Ravi et al. 2020; Guney et al. 2021; Pan et al. 2022; Chen et al. 2023).”
InfluentialState-Based Decoding of Force Signals From Multi-Channel Local Field Potentials
Ahmadi, A., Khorasani, A., Shalchyan, V., & Daliri, M. R. (2020) · IEEE Access · doi:10.1109/access.2020.3019267 · Cited by 22
A survey of deep learning-based classification methods for steady-state visual evoked potentials
Pan, Y., Chen, J., & Zhang, Y. (2023) · Brain-Apparatus Communication A Journal of Bacomics · doi:10.1080/27706710.2023.2181102 · Cited by 24
“Although the DL methods surpass the conventional frequency recognition methods, such as FBCCA and TRCA [31,48], in both UD and UI classification scenarios when a small amount of individual calibration data is available, the models with short data length still cannot achieve considerable…”
InfluentialLuo, W., Yin, W., Liu, Q., & Qu, Y. (2023) · Brain-Apparatus Communication A Journal of Bacomics · doi:10.1080/27706710.2023.2258938 · Cited by 24
SSVEP-DAN: Cross-Domain Data Alignment for SSVEP-Based Brain–Computer Interfaces
Chen, S.-Y., Chang, C.-M., Chiang, K.-J., & Wei, C.-S. (2024) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2024.3404432 · Cited by 21
“[38], [39].”
Human attention detection system using deep learning and brain–computer interface
Nair, S. A. L., & Megalingam, R. K. (2024) · Neural Computing and Applications · doi:10.1007/s00521-024-09628-8 · Cited by 20
An, Z., Wang, F., Wen, Y., Hu, F., & Han, S. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.124706 · Cited by 20
Zhang, X., Zhang, T., Jiang, Y., Zhang, W., Lu, Z., Wang, Y., & Tao, Q. (2024) · Heliyon · doi:10.1016/j.heliyon.2024.e26521 · Cited by 18
“Many published papers investigate various BCI system paradigms [2,64] and recognition algorithms [65,66].”
BackgroundSubject-specific CNN model with parameter-based transfer learning for SSVEP detection
Ji, Z., Xu, T., Chen, C., Yin, H., Wan, F., & Wang, H. (2024) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2024.107404 · Cited by 17
TSMNet: A comprehensive network based on spatio-temporal representations for SSVEP classification
Deng, L., Li, P., Zhang, H., Zheng, Q., Liu, S., Ding, X., Wang, M., & Gao, D. (2025) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2025.107554 · Cited by 17
Intelligent Classification Technique of Hand Motor Imagery Using EEG Beta Rebound Follow-Up Pattern
Wang, J., Chen, Y.-H., Yang, J., & Sawan, M. (2022) · Biosensors · doi:10.3390/bios12060384 · Cited by 16
“We adopted MI, although successful cases of other paradigms have been proposed, such as P300 ERP [42], SSVEP [43], spatial attention [44], selective attention [45], mental arithmetic [46], action observation [47], late positive potential (LPP) [48], etc.”
MethodGet a New Perspective on EEG: Convolutional Neural Network Encoders for Parametric t-SNE
Svantesson, M., Olausson, H., Eklund, A., & Thordstein, M. (2023) · Brain Sciences · doi:10.3390/brainsci13030453 · Cited by 16
“Li et al. used wavelet-generated features for parametric t-SNE [15], and Xu et al. extended this by combining wavelet features with the original EEG and frequency band features, all processed further through common spatial pattern filtering [16].”
BackgroundMethodA Zero-Padding Frequency Domain Convolutional Neural Network for SSVEP Classification
Gao, D., Zheng, W., Wang, M., Wang, L., Xiao, Y. J., & Zhang, Y. (2022) · Frontiers in Human Neuroscience · doi:10.3389/fnhum.2022.815163 · Cited by 15
“The deep learning method has been widely used in EEG, EMG, and other signals (Waytowich et al., 2018; Ravi et al., 2020; Zhang et al., 2021).”
MethodA CNN-Based Deep Learning Approach for SSVEP Detection Targeting Binaural Ear-EEG
Israsena, P., & Pan-ngum, S. (2022) · Frontiers in Computational Neuroscience · doi:10.3389/fncom.2022.868642 · Cited by 16
“Dataset also matters, as (Nakanishi et al., 2015; Ravi et al., 2020) already shown effects of different datasets used on performance evaluation.”
BackgroundMethodA TFA-CNN method for quantitative analysis in infrared spectroscopy
Ma, X., Zhang, S., Sun, J., Han, Y., Du, J., Fu, X., Yang, Y., Sa, Y., Li, Q., & Yang, C. (2022) · Infrared Physics & Technology · doi:10.1016/j.infrared.2022.104329 · Cited by 13
An Analysis of Traditional Methods and Deep Learning Methods in SSVEP-Based BCI: A Survey
Wu, J., & Wang, J. (2024) · Electronics · doi:10.3390/electronics13142767 · Cited by 12
Enhancing SSVEP-Based BCI Performance via Consensus Information Transfer Among Subjects
Zhang, X., Wei, W., Qiu, S., Li, X., Wang, Y., & He, H. (2024) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2024.3506998 · Cited by 12
“Another important approach is to abandon the template-matching scheme and instead design effective architectures for direct feature extraction and classification of SSVEP signals [26], [27], [28], [29], [30].”
A hybrid complex-valued neural network framework with applications to electroencephalogram (EEG)
Du, H., Riddell, R. P., & Wang, X. (2023) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2023.104862 · Cited by 15
Du, J., Ke, Y., Liu, S., Chen, S., & Ming, D. (2024) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2024.106282 · Cited by 14
Ikeda, A., & Washizawa, Y. (2021) · Sensors · doi:10.3390/s21165309 · Cited by 12
“For the CCNN, we used the same network structure as in [29], and we applied DFT with zero-padding; then, a complex spectrum from 3 Hz to 35 Hz was input to the network in a similar manner as [29].”
MethodA Convolutional Neural Network for SSVEP Identification by Using a Few-Channel EEG
Li, X., Yang, S., Fei, N., Wang, J., Huang, W., & Hu, Y. S. (2024) · Bioengineering · doi:10.3390/bioengineering11060613 · Cited by 11
“This input is consistent in form with previous studies [25,27,35,36].”
BackgroundMethodResultInfluentialFBDNN: filter banks and deep neural networks for portable and fast brain-computer interfaces
Bassi, P. R. A. S., & Attux, R. (2022) · Biomedical Physics & Engineering Express · doi:10.1088/2057-1976/ac6300 · Cited by 11
“Inspired by the results shown in [13], the FBCNN-2D is a two-dimensional convolutional neural network that analyzes complex spectrum”
InfluentialClassification of SSVEP Signals using Neural Networks for BCI Applications
Kumar, R. P., Vandana, S. S., Tejaswi, D., Charan, K., Janapati, R., & Desai, U. (2022) · International Conference on Intelligent Controller and Computing for Smart Power (ICICCSP) · doi:10.1109/iciccsp53532.2022.9862368 · Cited by 12
“The diverse training needs of feature extraction techniques for SSVEP-based BCI are analyzed and compared in this survey [9].”
MethodShishavan, H. H., Behzadi, M. M., Lohan, D. J., Dede, E. M., & Kim, I.-S. (2024) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2023.3345855 · Cited by 11
“We also compared the proposed model to various deep learning-based classifiers, selecting two deep-learning models for comparison - EEGNet [17] and UI-C-CNN [18], known for their high SSVEP detecting accuracy in both time and frequency domains.”
BackgroundMethodResultInfluentialChen, R., Xu, G., Zhang, H., Zhang, X., Xie, J., Tian, P., Zhang, S., & Han, C. (2023) · NeuroImage · doi:10.1016/j.neuroimage.2023.120501 · Cited by 10
“5, each method demonstrates dependable outcomes when contrasted with the findings of the preceding study (Ravi et al., 2020), and FBCCA surpasses CCA ( p 0.03) in the time frame of 0.5 s to 2.5 s.”
ResultInfluentialXu, D., Tang, F., Li, Y., Zhang, Q., & Feng, X. (2023) · Brain Sciences · doi:10.3390/brainsci13050780 · Cited by 10
“The FB-CCNN had a significantly higher accuracy of classification and stability compared to CCA and C-CNN.”
MethodInfluentialPhase-Locked Time-Shift Data Augmentation Method for SSVEP Brain-Computer Interfaces
Mai, X., Ai, J., Wei, Y., Zhu, X., & Meng, J. (2023) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2023.3323351 · Cited by 10
“However, low classification performance was obtained in previous SSVEP studies [36], [37] when time-shift augmentation method was directly used to augment more data epochs for training the spatial filtering algorithm.”
BackgroundMethodXie, S., Gao, W., Zeng, Z., Wu, Q. M. J., Huang, Q., Ban, N., Wu, Q., & Pan, J. (2024) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2024.108294 · Cited by 10
Wahab, A., Khan, U. S., Nawaz, T. H., Akbar, H., Hussain, S. T., Khalid, A., Ansari, A. R., & Nawaz, R. (2024) · IEEE Access · doi:10.1109/access.2024.3442235 · Cited by 10
“Usually, shorter time windows lead to lower accuracy but higher ITR and vice versa [30].”
BackgroundMethodDing, W., Liu, A., Xie, C., Wang, K., & Chen, X. (2024) · IEEE Transactions on Biomedical Engineering · doi:10.1109/tbme.2024.3458389 · Cited by 10
Xiao, X., Xu, L., Jin, Y., Pan, B., Xu, M., & Ming, D. (2022) · Journal of Neural Engineering · doi:10.1088/1741-2552/ac9861 · Cited by 11
“spectrum features [27], or lacked the comparison with state-of-the-art decomposition methods [26].”
BackgroundMethodLeon, M. S. I., Akter, J., Sakib, N., & Islam, M. K. (2021) · Lecture notes on data engineering and communications technologies · doi:10.1007/978-981-16-6636-0_45 · Cited by 9
A Calibration-Free Hybrid Approach Combining SSVEP and EOG for Continuous Control
Mai, X., Sheng, X., Shu, X., Ding, Y., Zhu, X., & Meng, J. (2023) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2023.3307814 · Cited by 10
“TDCA had insufficient recognition performance in continuous SSVEP-BCI because the extracted EEG segments are not phase-locked to each other [19], [20] (see details in Section I).”
BackgroundAn optimized Impulse Factor-based VMD-EMD approach to improve SSVEP accuracy for BCI Systems
Wahab, A., Khan, U. S., Nawaz, T. H., Hussain, S. T., Izhar, U., & Zeb, A. (2025) · Results in Engineering · doi:10.1016/j.rineng.2025.104203 · Cited by 9
A time-frequency feature fusion-based deep learning network for SSVEP frequency recognition
Dai, Y., Chen, Z., Cao, T., Zhou, H., Fang, M., Dai, Y., Jiang, L., & Tong, J. (2025) · Frontiers in Neuroscience · doi:10.3389/fnins.2025.1679451 · Cited by 9
Mahmood, S., Shin, J., Farhana, I., Islam, M. R., & Molla, M. K. I. (2021) · IEEE Access · doi:10.1109/access.2021.3136774 · Cited by 9
“The convolutional neural network (CNN) is employed with user-dependent (UD) and complex spectrum features for the detection of SSVEP [38].”
MethodLi, C., Liu, S., Wang, Z., & Yuan, G. (2023) · Frontiers in Physiology · doi:10.3389/fphys.2022.1085530 · Cited by 8
“Some studies have demonstrated that CV-CNN outperforms real-valued CNN after making full use of phase information in complex-valued data, such as magnetic resonance imaging (MRI) (Cole et al., 2021), steady-state visually evoked potentials (SSVEP) (Ravi et al., 2020).”
Ma, Y., Huang, J., Liu, C., & Shi, M. (2025) · Frontiers in Neurorobotics · doi:10.3389/fnbot.2024.1502560 · Cited by 8
Huang, J., Lv, Y., Zhang, Z., Xiong, B., Wang, Q., Wan, B., & Yang, P. (2024) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2024.3419013 · Cited by 8
“Previous time-shifting methods set the time-shift step to a fixed length of 50 or 100 ms [34], [35], which violates the time-locked feature of SSVEPs, resulting in low SSVEP decoding performance.”
MethodGao, Y., Ravi, A., & Jiang, N. (2021) · IEEE Access · doi:10.1109/access.2021.3112218 · Cited by 7
“Classification tools that are independent of the coefficient vectors, such as artificial neural networks [30], [32], can be used in”
BackgroundMethodSelf-citationA comparative survey of SSVEP recognition algorithms based on template matching of training trials
Luo, T.-J. (2022) · International Journal of Intelligent Computing and Cybernetics · doi:10.1108/ijicc-01-2022-0002 · Cited by 7
Multifunctional robot based on multimodal brain-machine interface
Ban, N., Xie, S., Qu, C., Chen, X., & Pan, J. (2024) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2024.106063 · Cited by 7
Bekhelifi, O., Berrached, N.-E., & Bendahmane, A. (2024) · Biomedical Physics & Engineering Express · doi:10.1088/2057-1976/ad2f58 · Cited by 8
Lan, Z., Li, Z., Yan, C., Xiang, X., Tang, D., Wu, M., & Chen, Z. (2025) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2025.3573410 · Cited by 7
“The complex convolutional neural network (CCNN) [19] shows remarkable performance in decoding SSVEP signals by taking the complex spectrum of SSVEP signals as input to the convolutional neural network (CNN).”
Decoding SSVEP Via Calibration-Free TFA-Net: A Novel Network Using Time-Frequency Features
Xu, L., Jiang, X., Wang, R., Lin, P., Yang, Y., Leng, Y., Zheng, W., & Ge, S. (2024) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2024.3510740 · Cited by 6
“C-CNN: Ravi et al. concatenate the real and imaginary components obtained by an FFT in the frequency dimension to form a feature vector as the input of a C-CNN [17].”
InfluentialMa, R., Cao, Y. K., Xie, S. Q., Zhang, M., Li, J., & Zhang, Z. (2025) · Neurocomputing · doi:10.1016/j.neucom.2025.131561 · Cited by 6
Ho, J., Wang, C., Chuang, C., King, C., Feng, C.-W., Chou, T.-H., Chen, Y.-M., Yang, Y.-H., & Hsiao, Y.-C. (2022) · IEEE International Conference on Human-Machine Systems (ICHMS) · doi:10.1109/ichms56717.2022.9980765 · Cited by 6
“In addition, the multi-target classification in [10] and the user-dependent or independent study in [8] have demonstrated that SSVEP-assisted BCI can be a helpful and practical method for many real-world applications.”
MethodWei, R., Hua, C., Chen, J., Mu, D., & Zhao, J. (2025) · IEEE Transactions on Systems Man and Cybernetics Systems · doi:10.1109/tsmc.2025.3578533 · Cited by 6
“1) CCNN [31]: CCNN uses frequency domain data rich in amplitude and phase information as input, extracting the frequency features of the signal through frequency domain convolution operations.”
Jiang, X., Cao, B., Ou, L. S., Chang, Y.-C., Do, T., & Lin, C.-T. (2025) · IEEE Transactions on Systems Man and Cybernetics Systems · doi:10.1109/tsmc.2025.3614244 · Cited by 10
“Consequently, supervised methods such as extended CCA (eCCA) [14], task-related component analysis (TRCA) [15], and complex-spectrum convolutional neural networks (CCNNs) [16] have been developed, significantly outperforming unsupervised approaches.”
Ravi, A., Wolfe, P., Tung, J. Y., & Jiang, N. (2025) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2025.3544479 · Cited by 5
“This paradigm does not require lengthy user training and little-to-no algorithm calibration, in contrast with BCIs based on motor imagery (MI), commonly investigated in BCI studies for stroke rehabilitation [14].”
Self-citationBinocularly incongruent, multifrequency-coded SSVEP in VR: feasibility and characteristics
Yang, L., Sun, Q., & Hulle, M. M. V. (2024) · Journal of Neural Engineering · doi:10.1088/1741-2552/ad775f · Cited by 5
“A similar issue happens in the study of [51] where the authors attribute this to the asynchronous training procedure.”
Abbasi, M. A., Abbasi, H. F., Yu, X., Aziz, M. Z., Yih, N. T. J. Y., & Fan, Z. (2024) · Journal of Neural Engineering · doi:10.1088/1741-2552/ad83f4 · Cited by 5
“Ravi et al [17] trains CNN on two types of features i.e. magnitude spectrum features and complex spectrum features and achieves 92.3% and 92.5% accuracy on two benchmark BCI datasets.”
InfluentialNeural–Computer Interfaces: Theory, Practice, Perspectives
Dubynin, I. A., Zemlyanskov, M. S., Shalayeva, I., Gorskii, O., Grinevich, V. P., & Евгеньевич, М. П. (2025) · Applied Sciences · doi:10.3390/app15168900 · Cited by 7
Training the spatially-coded SSVEP BCI on the fly
Maÿe, A., Mutz, M., & Engel, A. K. (2022) · Journal of Neuroscience Methods · doi:10.1016/j.jneumeth.2022.109652 · Cited by 4
“If this transfer captures the most relevant properties of the brain activity across a cohort of users, the effort for adapting the BCI to a new operator can be minimized [10].”
InfluentialA Multiple Command UAV Control System Based on a Hybrid Brain-Computer Interface
Wu, Q. M. J., Xie, S., Zeng, Z., Huang, Q., & Pan, J. (2023) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn54540.2023.10191621 · Cited by 4
“The SSVEP paradigm was often used in high-performance BCIs due to its high accuracy [10]–[12].”
MethodInfluentialAghili, S. N., Kilani, S., Rouhani, E., & Akhavan, A. (2023) · IEEE Access · doi:10.1109/access.2023.3347336 · Cited by 4
“Recently, CNN has attracted attention in SSVEP-based EEG [41], [49], [50], [51] and demonstrated higher performance in comparison with other traditional classifier methods [39], [40], [41] such as the CCA (when used as the classifier [3], [52]), neural network (NN), k-nearest neighbor (K-NN),…”
MethodData augmentation using masked principal component representation for deep learning-based SSVEP-BCIs
Ding, W., Liu, A., Cheng, L., & Chen, X. (2025) · Journal of Neural Engineering · doi:10.1088/1741-2552/add9d1 · Cited by 4
“Several deep learning models, such as EEGNet [15], fast Fourier transform-based convolutional neural network (FFT-CNN) [16], deep neural network (DNN) [17], time-domain-based CNN (tCNN) [18], SSVEP former [19], generalized zero-shot learning [20], and CNN integrated with transformer (CNN-Former)…”
Filter banks guided correlational convolutional neural network for SSVEPs based BCI classification
Wen, X., Jia, S., Han, D., Dong, Y., Gao, C., Cao, R., Hao, Y., Guo, Y., & Cao, R. (2024) · Journal of Neural Engineering · doi:10.1088/1741-2552/ad7f89 · Cited by 4
“Because short data lengths of EEG signals are necessary for individually calibrated SSVEP-BCIs [35], the EEG data were selected according to various time windows ( N l varying from 0.2 s to 1 s, with 0.2 s intervals).”
InfluentialRMKD: Relaxed matching knowledge distillation for short-length SSVEP-based brain–computer interfaces
Lan, Z., Li, Z., Yan, C., Xiang, X., Tang, D., Wu, M., & Chen, Z. (2025) · Neural Networks · doi:10.1016/j.neunet.2025.107133 · Cited by 4
Zhang, Y., Xia, M., Chen, K. G., Xu, P., & Yao, D. (2022) · PubMed · doi:10.7507/1001-5515.202102031 · Cited by 8
Li, P., Su, J., Belkacem, A. N., Cheng, L., & Chen, C. (2022) · Frontiers in Neuroscience · doi:10.3389/fnins.2022.971039 · Cited by 4
“At present, the SSVEP EEG signal classification method converts the original EEG signal through FFT and then inputs it into CNN for classification (Cecotti, 2011; Zhang et al., 2019; Ravi et al., 2020).”
Liu, Y., Song, Z., Xu, G., Wang, Z., Wan, F., Hu, Y. S., Zhang, M., & Zhang, Z. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10890554 · Cited by 3
“CCNN [21] focuses on spectral data using CNN and FFT, while SSVEPformer [22] combines CNN and MLP for spectrum-based feature encoding and attention learning.”
Bekhelifi, O., & Berrached, N.-E. (2024) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas58744.2024.10558617 · Cited by 3
“For SSVEP data the study [13] introduced a 2 convolutional blocks CNN that has the concatenation of amplitude and phase calculated by FFT as feature vector inputs.”
MethodAn improved BCI-VR hybrid interactive system based on a mixed template CCA method
Niu, L., An, L., Bin, J., Zhao, Y., Kang, X., Zhang, L., & Ge, H. (2025) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2025.107907 · Cited by 3
Wang, L., Zhang, Z., Han, D., Zhang, Z., Liu, Z., & Liu, W. (2020) · European Journal of Neuroscience · doi:10.1111/ejn.15030 · Cited by 3
Asynchronous SSMVEP BCI and Influence of Dynamic Background in Augmented Reality
Ravi, A., Lu, J., Pearce, S., & Jiang, N. (2021) · International Conference on Mechatronics and Machine Vision in Practice (M2VIP) · doi:10.1109/m2vip49856.2021.9665078 · Cited by 3
“More recently, by providing the complex spectrum as input to a CNN (C-CNN), superior asynchronous performance over FFT-CNN was demonstrated in both user-dependent and user-independent training for SSVEP BCI [15].”
BackgroundMethodInfluentialSelf-citationKoraganji, V. N., Whelan, A. J., Mokkapati, A., Zerick, J. N., Winters, R. M., & Lewis, G. F. (2021) · Conference proceedings/Conference proceedings - IEEE International Conference on Systems, Man, and Cybernetics · doi:10.1109/smc52423.2021.9658606 · Cited by 3
“The choice of window size for CNN architecture comparison was a difficult one, but we settled on 1s based on its overall performance in various studies using CNNs for EEG data [48], [35], [34].”
MethodKim, J., Kim, T., & Lee, J. (2024) · CHI Conference Extended Abstracts on Human Factors in Computing Systems (CHI EA) · doi:10.1145/3613905.3651084 · Cited by 2
“In addition to user-dependent (Intra-subject) evaluation, we evaluated independent (Inter-subject) cases to explore the generalization and possibility of training-free usage [39, 47].”
Metadata
- Authors
- Ravi, A., Beni, N. H., Manuel, J., & Jiang, N.
- Affiliations
- University of Waterloo (CA)
- Corresponding author
- Jiang, Ning
- Venue
- Journal of Neural Engineering
- Publisher
- IOP Publishing
- Published
- 2020-04-01
- Volume
- 17
- Issue
- 2
- Pages
- 026028
- ISSN
- 1741-2552 (ISSN-L)
- 1741-2560
- OpenAlex
- W3000563309
- Cited by
- 165
- References
- 46
- Citations by year
- 2026: 14 · 2025: 43 · 2024: 40 · 2023: 22 · 2022: 24 · 2021: 17 · 2020: 5
- FWCI
- 9.15
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- creativecommons.org/licenses/by/3.0/ (vor, 7 d)
- iopscience.iop.org/info/page/text-and-data-mining (tdm, 7 d)
- Locations
- Journal of Neural Engineering (publishedVersion, cc-by)
- PubMed (publishedVersion)
- Funders
- Natural Sciences and Engineering Research Council of Canada (401261605)
- NSERC – CREATE, Training in Global Biomedical Technology Research and Innovation at the University of Waterloo. (401207296)
- Topics
- Life Sciences › Neuroscience › Cognitive Neuroscience › EEG and Brain-Computer Interfaces
- Physical Sciences › Computer Science › Human-Computer Interaction › Gaze Tracking and Assistive Technology
- Physical Sciences › Computer Science › Artificial Intelligence › Emotion and Mood Recognition
- Keywords
- steady-state visual evoked potentials, SSVEP-based brain-computer interface, convolutional neural networks, canonical correlation analysis, filter bank canonical correlation analysis, SSVEP classification, EEG feature extraction
- MeSH terms
- Algorithms, Electroencephalography, Evoked Potentials, Evoked Potentials, Visual, Humans, Photic Stimulation, Neural Networks, Computer, Brain-Computer Interfaces
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.