CoChat

Comparing user-dependent and user-independent training of CNN for SSVEP BCI

Open access
Ravi, Aravind, Beni, Nargess Heydari, Manuel, Jacob, Jiang, Ning · 2020 · Journal of Neural Engineering

The 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.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1088/1741-2552/ab6a67

Cites

47 of 47 references verified0 citation errors · 0 not checkable

Statistics

Figures

Image 1
CC-BY · From the PDF
Image 2
CC-BY · From the PDF
Figure 3 Performance comparison of Dataset 1 -(a) comparison of the average accuracies, (b) comparison of the average ITR (bits/min), across all participants for the different classification methods for data lengths of W = { 0.5 s, 1 s, 1.5 s, 2 s, 2.5 s, 3 s } . Chance level = 0.14. The vertical bars indicate the standard deviation among the participants at each W .
Figure 3 Performance comparison of Dataset 1 -(a) comparison of the average accuracies, (b) comparison of the average ITR (bits/min), across all participants for the different classification methods for data lengths of W = { 0.5 s, 1 s, 1.5 s, 2 s, 2.5 s, 3 s } . Chance level = 0.14. The vertical bars indicate the standard deviation among the participants at each W .CC-BY · From the PDF
Figure 4 The performance comparison of Dataset 2 -(a) comparison of the average accuracies, (b) comparison of the average ITR (bits/min) across all participants for the different classification methods for data lengths of W = 1 s. Chance level = 0.083. The vertical line overlaying each bar indicates the standard deviation among all participants in each method.
Figure 4 The performance comparison of Dataset 2 -(a) comparison of the average accuracies, (b) comparison of the average ITR (bits/min) across all participants for the different classification methods for data lengths of W = 1 s. Chance level = 0.083. The vertical line overlaying each bar indicates the standard deviation among all participants in each method.CC-BY · From the PDF
Figure 5 Dataset 1 -(a) -(c) feature visualization of an unseen participant using t-SNE -UI-M-CNN. (d) -(f) Feature visualization of an unseen participant using t-SNE -UI-C-CNN. (a) Input magnitude spectrum features. (b) Output of Conv_1_ ReLU Layer of M- CNN. (d) Input complex spectrum features. (e) Output of Conv_1_ReLU Layer of C-CNN. (f) Output of the Conv_2_ReLU layer of C-CNN.
Figure 5 Dataset 1 -(a) -(c) feature visualization of an unseen participant using t-SNE -UI-M-CNN. (d) -(f) Feature visualization of an unseen participant using t-SNE -UI-C-CNN. (a) Input magnitude spectrum features. (b) Output of Conv_1_ ReLU Layer of M- CNN. (d) Input complex spectrum features. (e) Output of Conv_1_ReLU Layer of C-CNN. (f) Output of the Conv_2_ReLU layer of C-CNN.CC-BY · From the PDF
Figure 6 Dataset 2 -feature visualization of all participants using t-SNE. (a) Output of the Conv_2_ReLU layer of UI-M-CNN. (b) Output of the Conv_2_ReLU layer of UI-C-CNN (Right), (c) segment level clustering for SSVEP Class 12.25 Hz of UI-C-CNN.
Figure 6 Dataset 2 -feature visualization of all participants using t-SNE. (a) Output of the Conv_2_ReLU layer of UI-M-CNN. (b) Output of the Conv_2_ReLU layer of UI-C-CNN (Right), (c) segment level clustering for SSVEP Class 12.25 Hz of UI-C-CNN.CC-BY · From the PDF
Image 7
CC-BY · From the PDF

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.

A few free checks a day · papers already verified open instantly · results are public

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

Cited by

See all 170

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
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
Licence
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.