English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Auditory Attention Detection (AAD) aims to detect target speaker from brainsignals in a multi-speaker environment. Although EEG-based AAD methods haveshown promising results in recent years, current approaches primarily rely ontraditional convolutional neural network designed for processing Euclidean datalike images. This makes it challenging to handle EEG signals, which possessnon-Euclidean characteristics. In order to address this problem, this paperproposes a dynamical graph self-distillation (DGSD) approach for AAD, whichdoes not require speech stimuli as input. Specifically, to effectivelyrepresent the non-Euclidean properties of EEG signals, dynamical graphconvolutional networks are applied to represent the graph structure of EEGsignals, which can also extract crucial features related to auditory spatialattention in EEG signals. In addition, to further improve AAD detectionperformance, self-distillation, consisting of feature distillation andhierarchical distillation strategies at each layer, is integrated. Thesestrategies leverage features and classification results from the deepestnetwork layers to guide the learning of shallow layers. Our experiments areconducted on two publicly available datasets, KUL and DTU. Under a 1-secondtime window, we achieve results of 90.0\% and 79.6\% accuracy on KUL and DTU,respectively. We compare our DGSD method with competitive baselines, and theexperimental results indicate that the detection performance of our proposedDGSD method is not only superior to the best reproducible baseline but alsosignificantly reduces the number of trainable parameters by approximately 100times.

DGSD: Dynamical Graph Self-Distillation for EEG-Based Auditory Spatial Attention Detection