Skip to content
  • Projects
  • Publications
  • News
  • About us
    • Our team
    • PhD students
    • Eriksholm’s timeline
    • History of hearing aids
    • Partnerships
    • Contact us
  • Projects
  • Publications
  • News
  • About us
    • Our team
    • PhD students
    • Eriksholm’s timeline
    • History of hearing aids
    • Partnerships
    • Contact us
  • Projects
  • Publications
  • News
  • About us
    • Our team
    • PhD students
    • Eriksholm’s timeline
    • History of hearing aids
    • Partnerships
    • Contact us
  • Projects
  • Publications
  • News
  • About us
    • Our team
    • PhD students
    • Eriksholm’s timeline
    • History of hearing aids
    • Partnerships
    • Contact us
Back

Deep Neural Networks for speaker separation and speech enhancement 

Lars Bramsløw

Principal Scientist

Eriksholm Research Centre

Deep Neural Networks

Introduction

The healthy ear is fantastic and allows us to focus on particular talkers and messages in even very noisy conditions. With hearing loss, this ability is reduced, and hearing aid users often complain about hearing out particular voices in a crowd or in generally noisy conditions such as party. Deep neural networks (DNN) can be trained to separate voices from other voices or from noise and hence reduce the listening challenges in difficult listening situations. This project investigates the potential benefit for hearing aid users.

Funded by the William Demant Foundation.

Aims

In collaboration with Tampere University, the project aims at designing, training and evaluating DNN speaker separation with the aim of improving speech intelligibility in competing-voice scenarios as well as speech-in-noise scenarios.

Ideal and estimated ratio mask

Methodology

After selection of the most relevant scenarios, training materials were assembled from own and public speech and noise corpora and used for training the DNN system. Other excerpts from the same corpora were saved and used for testing the separation performance, first with objective signal measures and afterwards in listening tests with hearing impaired listeners. In the competing-voice scenario, the two separated outputs were fed to the two ears, and in the speech-in-noise scenario, the separated speech was fed to both ears.

Separation demo

Results

The listening tests documented significant benefit for hearing impaired listeners –

1) in the presence of one competing talker, a 13%-point increase in speech intelligibility was measured when listening to the target and competing voice in each ear

2) in the presence of one competing talker, a 37%-point benefit was registered when listening to the target voice in both ears (from 58% to 95%)

3) a 16%-point benefit was registered when listening to speech separated from party noise

Publications

Loading...
Naithani, G., Barker, T., Parascandolo, G., Bramsløw, L., Pontoppidan, N., Virtanen, T. (2017). Low latency sound source separation using convolutional recurrent neural networks [Conference Proceedings]. the 2017 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, Mohonk, USA.
Bramsløw, L., Naithani, G., Hafez, A., Barker, T., Pontoppidan, N. H., & Virtanen, T. (2018). Improving competing voices segregation for hearing impaired listeners using a low-latency deep neural network algorithm. the Journal of the Acoustical Society of America/the Journal of the Acoustical Society of America, 144(1), 172–185. https://doi.org/10.1121/1.5045322
Bramsløw, L., Vatti, M., Rossing, R., Naithani, G., & Pontoppidan, N. H. (2019). A competing voices test for Hearing-Impaired listeners applied to spatial separation and ideal Time-Frequency masks. Trends in Hearing, 23, 233121651984828. https://doi.org/10.1177/2331216519848288
Beck, D. L., Bramsløw, L. (2019). Speech in Noise Research and Deep Neural Networks: An Interview with Lars Bramslow, PhD. Hearing Review. 2019;26(9):46-47

Team

Lars Bramsløw

Principal Scientist

Eriksholm Research Centre

Gaurav Naithani

PhD student

Tampere University

Tuomas Virtanen

Professor

Tampere University

Joonas Nikunen

PostDoc

Tampere University

View all

Partners

  • Tampere University, Finland

You may also be interested in

Loading...
Individualized hearing loss compensation via auditory models and deep neural networks
CURRENT
Individualized hearing loss compensation via auditory models and deep neural networks
7622,4459,7625

Personalised Audiology, Artificial Intelligence (AI)

Compensating for hearing loss by providing amplification specific to the hearing loss is the core fu...
Compensating for hearing loss by providing amplification specific to the hearing loss is the core…

You may also be interested in

Individualized hearing loss compensation via auditory models and deep neural networks
CURRENT
Individualized hearing loss compensation via auditory models and deep neural networks
7622,4459,7625

Personalised Audiology, Artificial Intelligence (AI)

Compensating for hearing loss by providing amplification specific to the hearing loss is the core fu...
Compensating for hearing loss by providing amplification specific to the hearing loss is the core…

People are our most
valuable source of
insights

Facebook

Instagram

LinkedIn

Youtube

  • Eriksholm Research Centre
  • Rørtangvej 20
  • DK-3070 Snekkersten
  • Denmark
We are a part of Oticon, a world leader in hearing care. We share the same philosophy that people are our main source of insights
Bliv testperson
  • +45 48 29 89 00
  • mail@eriksholm.com
  • Cookie policy
  • Disclaimer

© 2025 Eriksholm – Designed by Aveo web&marketing

Manage consent to cookies
We use cookies to optimize our website and our service.
Functional Always active
The technical storage or access is strictly necessary for the legitimate purpose of enabling the use of a specific service expressly requested by the subscriber or user, or solely for the purpose of transmitting a communication via an electronic communication network.
Preferences
The technical storage or access is necessary for the legitimate purpose of storing preferences that are not requested by the subscriber or user.
Statistics
The technical storage or access used exclusively for statistical purposes. The technical storage or access that is used exclusively for anonymous statistical purposes. Without a subpoena, voluntary compliance on the part of your Internet Service Provider, or additional records from a third party, information stored or retrieved for this purpose alone cannot usually be used to identify you.
Marketing
The technical storage or access is necessary to create user profiles for the purpose of sending advertisements or to track the user on a website or across multiple websites for similar marketing purposes.
Manage options Manage services Manage {vendor_count} vendors Read more about these purposes
Preferences
{title} {title} {title}