P²CSL: cross-subject EEG classification by subspace class prototype-based progressive confident target sample labeling.

Abstract

<i>Objective.</i>Domain adaptation (DA) has achieved remarkable performance in cross-subject electroencephalogram (EEG) decoding by mitigating the inter-subject data distribution discrepancies. However, when exploring the feature alignment subspace and performing self-supervised pseudo-labeling in an iterative way, two difficulties are often encountered: one is that unreliable target labeling results inevitably mislead the domain-free feature learning process in the early stage and the other is that the contribution of source and target samples should be balanced in the later stage.<i>Approach.</i>To address both issues, this paper proposes prototype-based progressive confident target sample labeling (P²CSL) method to use subspace class prototypes to assist in labeling target samples under the unified framework of domain-invariant EEG feature learning and the self-supervised target sample labeling, and progressively incorporate confident target samples into DA model fitting. The underlying rationality is that early-stage pseudo-labels from unconverged models are prone to error propagation, requiring auxiliary mechanisms to ensure their reliability and stabilize training. With the gradual alignment of cross-subject features, the estimated pseudo-label information of target domain will be more reliable, meaning that more target samples should be involved in model training.<i>Main results.</i>Experiments on emotion recognition and inner speech decoding demonstrate the competitive performance of P²CSL in cross-subject EEG classification in comparison with SOTA methods.<i>Significance.</i>Our study indicates the effectiveness of jointly considering the reliability of target samples and their contribution to model training in the context of DA. In addition, some fine-grained results including the sample confidence allocation strategy, the DA effects, and the dynamic model optimization process are provided to further illustrate the model execution details.