End-to-End Deep Neural Network for Auditory Attention Decoding

The ability to analyze auditory attention has a range of applications, from improving hearing aids to enhancing the understanding of how our brains process sound. This ability is crucial in various sectors, including healthcare, artificial intelligence, and consumer electronics. There exists a need for a more sophisticated method to identify what a listener is paying attention to amidst multiple acoustic sources. Current methods for analyzing auditory attention often rely solely on acoustic analysis. This approach fails to account for the individual listener's focus, making it difficult to accurately identify the sound source to which they are attentive. The challenge lies in accurately distinguishing between the multiple sound sources and determining which one the listener is focused on.

Technology Description

This technology involves receiving and analyzing neural data in response to a listener's auditory attention. It also involves receiving an acoustic signal responsive to several acoustic sources. For each of these sources, the method generates audio data containing the characteristics of the acoustic source and forms combined data representative of both the neural data and the audio data. This combined data is then provided to a classification network that calculates a similarity score between the neural data and the acoustic source using one or more similarity metrics. This method is unique because it identifies an acoustic source associated with a listener's auditory focus using similarity scores. These scores are calculated for each of the multiple acoustic sources. This innovative approach provides real-time analysis, differentiating between multiple sources to identify which is commanding the listener's attention at a given point. The integration of neural data alongside acoustic data gives the technology a significant edge in determining the focus of auditory attention.