PCRNet: Phase-aware Complex Refinement Network for EEG-based Auditory Attention Decoding

Auditory attention decoding (AAD) based on Electroencephalography (EEG) aims to identify the attended speaker in multi-speaker environments. However, existing methods typically overlook the crucial phase information of EEG signals, which limits their ability to distinguish structured neural patterns from random noise in the frequency domain and hinders robust decoding. To address these issues, this paper proposes a Phase-aware Complex Refinement Network (PCRNet) for AAD, which consists of a Temporal Context Calibration (TCC) module and a Dual-Domain Integration (DDI) module. Specifically, the TCC module captures long-range temporal dependencies through multi-scale temporal attention mechanism, while the DDI module employs a phase-guided spectral filtering strategy to dynamically suppress noise-dominated frequencies and refine the real and imaginary components separately. This design enables effective phase recalibration and enhances the discriminability of target features in the complex domain. Experimental results on three public datasets demonstrate that PCRNet outperforms state-of-the-art (SOTA) methods, particularly under challenging ultra-short 0.1-second windows. Code is available at: https://github.com/SunshineGreeny/PCRNet.

Imagine trying to follow one person’s voice in a crowded room. Humans can do this naturally, but it is difficult for computers to know which speaker a listener is focusing on. Our research was motivated by this challenge: how can we help computers infer a person’s listening focus from brain activity? This problem is important for future hearing-assistance systems that could automatically enhance the voice a user wants to hear. We study this problem using EEG signals, which record brain activity from the scalp. EEG signals are weak, noisy, and hard to interpret, so many existing methods mainly look at signal strength. However, they often miss an important timing clue called phase, which describes how waves in a signal are aligned over time. We propose PCRNet, a neural network that makes better use of this phase-related information. It learns useful timing patterns and refines frequency-domain information, helping separate meaningful brain responses from random noise. Across three public datasets, PCRNet performs better than previous methods, especially when only very short EEG segments are available. This matters because real hearing-assistance systems need to respond quickly and reliably. Our results suggest that using phase information can help build faster and more robust brain-guided hearing technology.