Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Spatial Distillation based Distribution Alignment (SDDA) for Cross-Headset EEG Classification

Dingkun Liu, Siyang Li, Ziwei Wang, Wei Li, Dongrui Wu

TL;DR

This paper tackles the challenge of cross-headset EEG classification where different headsets have varying electrode configurations. It introduces Spatial Distillation based Distribution Alignment (SDDA), a teacher–student framework in which a teacher trained on a full electrode set guides a student operating on the common channel subset, augmented by session-wise Euclidean alignment and multi-stage distribution alignment (MA and CL) to bridge input, feature, and output gaps. The approach yields improvements in both offline unsupervised and online supervised domain adaptation across six MOABB EEG datasets spanning motor imagery and P300 paradigms, consistently surpassing multiple state-of-the-art baselines. The work demonstrates the practical potential of leveraging extra channels via knowledge distillation for robust cross-headset BCI calibration, with code released for reproducibility and further research.

Abstract

A non-invasive brain-computer interface (BCI) enables direct interaction between the user and external devices, typically via electroencephalogram (EEG) signals. However, decoding EEG signals across different headsets remains a significant challenge due to differences in the number and locations of the electrodes. To address this challenge, we propose a spatial distillation based distribution alignment (SDDA) approach for heterogeneous cross-headset transfer in non-invasive BCIs. SDDA uses first spatial distillation to make use of the full set of electrodes, and then input/feature/output space distribution alignments to cope with the significant differences between the source and target domains. To our knowledge, this is the first work to use knowledge distillation in cross-headset transfers. Extensive experiments on six EEG datasets from two BCI paradigms demonstrated that SDDA achieved superior performance in both offline unsupervised domain adaptation and online supervised domain adaptation scenarios, consistently outperforming 10 classical and state-of-the-art transfer learning algorithms.

Spatial Distillation based Distribution Alignment (SDDA) for Cross-Headset EEG Classification

TL;DR

This paper tackles the challenge of cross-headset EEG classification where different headsets have varying electrode configurations. It introduces Spatial Distillation based Distribution Alignment (SDDA), a teacher–student framework in which a teacher trained on a full electrode set guides a student operating on the common channel subset, augmented by session-wise Euclidean alignment and multi-stage distribution alignment (MA and CL) to bridge input, feature, and output gaps. The approach yields improvements in both offline unsupervised and online supervised domain adaptation across six MOABB EEG datasets spanning motor imagery and P300 paradigms, consistently surpassing multiple state-of-the-art baselines. The work demonstrates the practical potential of leveraging extra channels via knowledge distillation for robust cross-headset BCI calibration, with code released for reproducibility and further research.

Abstract

A non-invasive brain-computer interface (BCI) enables direct interaction between the user and external devices, typically via electroencephalogram (EEG) signals. However, decoding EEG signals across different headsets remains a significant challenge due to differences in the number and locations of the electrodes. To address this challenge, we propose a spatial distillation based distribution alignment (SDDA) approach for heterogeneous cross-headset transfer in non-invasive BCIs. SDDA uses first spatial distillation to make use of the full set of electrodes, and then input/feature/output space distribution alignments to cope with the significant differences between the source and target domains. To our knowledge, this is the first work to use knowledge distillation in cross-headset transfers. Extensive experiments on six EEG datasets from two BCI paradigms demonstrated that SDDA achieved superior performance in both offline unsupervised domain adaptation and online supervised domain adaptation scenarios, consistently outperforming 10 classical and state-of-the-art transfer learning algorithms.

Paper Structure

This paper contains 20 sections, 12 equations, 5 figures, 6 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Transfer Learning
  4. Cross-Headset Transfer
  5. SDDA
  6. Problem Definition
  7. Spatial Distillation
  8. Distribution Alignment
  9. Session-wise EA
  10. Marginal Alignment (MA)
  11. Confusion Loss (CL)
  12. Summary
  13. Experiments
  14. Datasets
  15. Experiment Settings
  16. ...and 5 more sections

Figures (5)

  • Figure 1: Transfer learning for BCIs.
  • Figure 2: Architecture of the proposed SDDA for cross-headset EEG classification. The source data with a full set of electrodes are used to train the teacher model. The student model is trained on source data using common electrodes with the target domain. Target data are incorporated to align the probability distributions and reduce the prediction confusion. Cross-entropy loss is applied to the labeled source data and a small amount of labeled target calibration data.
  • Figure 3: Two different cross-headset transfer settings. (a) UDA; and, (b) SDA.
  • Figure 4: Ablation study results.
  • Figure 5: $t$-SNE visualization of the data in BNCI2014004. (a) Before EA; (b) After EA. Different colors represent trials from different subjects.