AFA‑Net introduces a differential attention mechanism to improve Auditory Attention Detection (AAD) from EEG signals, explicitly targeting noisy data. The framework outperforms existing deep learning models, achieving 96.8% accuracy within a 2‑second decision window while using fewer parameters. It represents one of the first approaches to actively mitigate EEG noise in AAD tasks.
arXiv:2607. 18345v1 Announce Type: cross Abstract: Limited training data constrains deep learning models for Auditory Attention Decoding (AAD) in hearing aids (HAs).
arXiv:2606. 14120v1 Announce Type: cross Abstract: Auditory attention decoding (AAD) aims to infer the attended speaker from neural responses in multi-speaker acoustic environments and is a key problem for neuro-steered hearing systems.
RAMamba-Net is a new multimodal fusion network designed for auditory attention decoding (AAD) that combines EEG and electrooculography (EOG) signals. It uses a Mamba-enhanced band-aware convolutional Transformer to capture EEG band-specific patterns and long-range temporal dynamics, while a dual-branch encoder models EOG temporal and inter-channel dependencies. Cross‑modal attention and a reliability‑aware module estimate sample‑wise modality weights, improving fusion robustness and achieving a 5.76% accuracy gain over unimodal baselines on two AAD benchmarks.
The paper introduces Adaptive Anisotropic Attention (AAA), a method that splits self‑attention into temporal and spatial paths for axis‑structured signals like EEG. A learned gate combines the two paths for each token, and the resulting AXON model outperforms dense attention baselines on six EEG tasks and shows transfer to audio spectrograms. The study demonstrates that aligning attention with the natural axes of structured data provides a beneficial inductive bias.
arXiv:2503. 00340v2 Announce Type: cross Abstract: Lightweight models are essential for real-time speech enhancement applications.
The paper introduces Adaptive Anisotropic Attention (AAA), a method that splits self‑attention into temporal and spatial paths for axis‑structured signals like EEG. A learned gate combines the two paths for each token, and the resulting AXON model outperforms dense attention baselines on six EEG tasks and shows benefits in audio spectrogram experiments. The study demonstrates that aligning attention with natural signal axes provides a useful inductive bias.
arXiv:2606. 12662v1 Announce Type: cross Abstract: Speech enhancement models typically apply uniform capacity across all frequencies, disregarding the non-uniform spectral resolution of human hearing.
The paper introduces EmoSpeechBrain, a multimodal emotion recognition framework that fuses EEG and speech signals. It employs differential attention in the EEG encoder to cancel shared noise and an attention-based gating adapter to align modalities and weight their contributions. Experiments on PME4 and EAV datasets show up to 12.9% accuracy improvement over other EEG encoders and surpass unimodal baselines by up to 23.1%.
arXiv:2606. 24164v1 Announce Type: cross Abstract: Recent end-to-end models for EEG-guided target speech extraction report impressive results, underscoring potential for neuro-steered hearing technologies.
SHINE is a Sequential Hierarchical Integration Network designed to reconstruct speech envelope and Mel spectrogram from EEG and MEG recordings. It uses a residual sensor adapter, dilated-block states for temporal depth, and a target- and time-dependent gate to fuse hierarchical and attention-enhanced context predictions. Across two EEG and two MEG datasets, SHINE achieved the highest mean envelope and mean-Mel Pearson correlations among nine baseline methods and ranked second in the NeurIPS 2025 PNPL Competition’s speech-detection Extended Track.
Recent end-to-end models for EEG-guided target speech extraction report impressive results, underscoring potential for neuro-steered hearing technologies. However, our analysis reveals that high within-trial performance can be driven by trial-specific EEG structure that acts as shortcuts for target selection, leading to poor generalization on unseen trials.
arXiv:2606. 11922v1 Announce Type: cross Abstract: Recent respiratory sound classification (RSC) studies largely rely on CLS-token driven self-attention architectures such as the Audio Spectrogram Transformer (AST).