Decoding Stimulus Reconstruction-Based Auditory Attention Robustly in Unbalanced EEG Datasets
For researchers using EEG-based auditory attention decoding, this work addresses a critical methodological gap by ensuring robust evaluation on unbalanced datasets, which are common in published studies.
The study reveals that stimulus reconstruction-based deep neural network decoders for auditory attention detection (AAD) produce overestimated performance on unbalanced EEG datasets. The proposed leave-one-paired-envelope-out (LOPEO) cross-validation protocol effectively prevents this inflation, providing a principled evaluation framework for unbalanced datasets.
In the past decade, numerous studies have applied deep neural networks (DNNs) to decode auditory attention (AAD) from Electroencephalogram (EEG) signals via stimulus reconstruction. However, the influence of dataset balance on the decoding performance of stimulus reconstruction-based AAD remains unexplored. In this study, three publicly available EEG-AAD datasets - KUL, DTU, and NJU cEEGrid - are used to construct both balanced and unbalanced experimental conditions. We hypothesize and demonstrate that stimulus reconstruction-based DNN decoders tend to produce overestimated decoding performance on unbalanced datasets. To address this issue, we propose a leave-one-paired-envelope-out (LOPEO) cross-validation protocol. Experimental results confirm that LOPEO effectively prevents inflated decoding accuracy on unbalanced datasets. While balanced datasets are generally preferred in experimental design, LOPEO provides a principled evaluation framework for unbalanced datasets that have already been published, filling an important gap in the field.