Excitation-inhibition Cell activity patterns for binaural source localisation
Excitation-inhibition Cell activity patterns for binaural source localisation
This paper introduces a novel approach to source localisation that combines existing models of the auditory periphery with a deep neural network model. The model is designed to mimic Excitation-Inhibition (EI) cell activity in the human auditory system that is believed to enable localisation. The paper investigates the performance of the proposed approach for a single source under anechoic conditions. The results exhibit the characteristics of human source localisation, including the phenomenon of front-back confusion. The results are benchmarked against a state-of-the-art deep learning technique that is based on a convolutional neural network.
binaural hearing, source localisation, deep neural networks
Wang, Hsuan-Yang
77ff593b-fbf9-4d92-8f42-03ed0b792531
Nelson, Philip
5c6f5cc9-ea52-4fe2-9edf-05d696b0c1a9
Evers, Christine
93090c84-e984-4cc3-9363-fbf3f3639c4b
Wang, Hsuan-Yang
77ff593b-fbf9-4d92-8f42-03ed0b792531
Nelson, Philip
5c6f5cc9-ea52-4fe2-9edf-05d696b0c1a9
Evers, Christine
93090c84-e984-4cc3-9363-fbf3f3639c4b
Wang, Hsuan-Yang, Nelson, Philip and Evers, Christine
(2021)
Excitation-inhibition Cell activity patterns for binaural source localisation.
IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, Mohonk Mountain House, New Paltz, NY, United States.
17 - 20 Oct 2021.
5 pp
.
(In Press)
Record type:
Conference or Workshop Item
(Paper)
Abstract
This paper introduces a novel approach to source localisation that combines existing models of the auditory periphery with a deep neural network model. The model is designed to mimic Excitation-Inhibition (EI) cell activity in the human auditory system that is believed to enable localisation. The paper investigates the performance of the proposed approach for a single source under anechoic conditions. The results exhibit the characteristics of human source localisation, including the phenomenon of front-back confusion. The results are benchmarked against a state-of-the-art deep learning technique that is based on a convolutional neural network.
More information
Accepted/In Press date: 14 July 2021
Venue - Dates:
IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, Mohonk Mountain House, New Paltz, NY, United States, 2021-10-17 - 2021-10-20
Keywords:
binaural hearing, source localisation, deep neural networks
Identifiers
Local EPrints ID: 450599
URI: http://eprints.soton.ac.uk/id/eprint/450599
PURE UUID: 167efec6-ed03-4fa7-b6fc-f9ec1b819031
Catalogue record
Date deposited: 04 Aug 2021 16:36
Last modified: 14 Mar 2024 03:17
Export record
Contributors
Author:
Hsuan-Yang Wang
Author:
Christine Evers
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics