Tool captures brain signals tracking speech in noisy settings
A multi-scenario EEG dataset for auditory attention decoding in naturalistic multi-talker environments
Sound
Summary
People’s brains can focus on one voice in noisy places, but studying this has been limited by simple setups. The authors created a large EEG dataset capturing brain activity as listeners focus on one speaker among many in realistic environments like restaurants and meetings. This helps researchers study how our brains follow speech in real life and can improve hearing devices that adapt to what a user wants to hear. The dataset records many channels of brain signals from 30 people over 25 hours in natural multi-speaker scenes.
What this means in practice
- •For hearing aid developers: Train and evaluate neuro-steered hearing aids using realistic brain data from multi-speaker scenarios to improve selective listening.$Commercial implications: The dataset enables building hearing aids that decode user attention in real-world noisy environments, improving device performance and user experience.
- •For audio signal processing teams: Benchmark and improve algorithms that separate important speech from background noise using realistic brain responses to complex acoustic scenes.
Authors
Shu Peng, Rui Liu, Yufei Zhang, Wenlong You, Zhige Chen, Jiachen Xi, Qiyuan Sun, Yan Liu, Kay Chen Tan, Jibin Wu
Abstract
Understanding how the brain selectively follows relevant speech amid competing voices is a central challenge in auditory neuroscience and a key step toward neuro-steered hearing technologies. However, most open-source Electroencephalography (EEG) datasets for Auditory Attention Decoding (AAD) use idealized single-competing-talker paradigms that oversimplify the acoustic, spatial, and semantic structure of everyday communication. To capture this ecological complexity, we introduce the SoundBubble-EEG dataset: a high-density 128-channel EEG resource comprising more than 25 hours of recordings from 30 participants. The paradigm requires listeners to selectively attend to a dynamic target speaker group, a designated "sound bubble", amid competing multi-speaker distractor bubbles across three realistic scenarios: a restaurant, a home TV viewing, and a meeting discussion. By bridging the gap between constrained laboratory protocols and real-world auditory scenes, this dataset enables investigations of multi-talker speech comprehension, neural speech tracking, and cross-scenario generalization. It also provides a benchmark for AAD algorithms under realistic acoustic and semantic variability and may support auditory neuroscience and the development of neuro-steered hearing technologies.