Research Associate in Machine Learning-Based Spatial Audio

Imperial College London Department of Electrical and Electronic Engineering

Job summary

We have a research associate (postdoc) position to work on spatial audio processing and spatial hearing using methods from machine learning. The aim of the project is to design a method for interactively fitting individualised filters for spatial audio (HRTFs) to users in real-time based on their interactions with a VR/AR environment. We will use meta-learning algorithms to minimise the time required to individualise the filters, using...

Job listing information

  • Reference ENG01810
  • Date posted 20 July 2021
  • Closing date 20 September 2021

Key information about the role

  • Location South Kensington Campus (map)
  • Position type Full time, fixed term
  • Salary £40,858 – £48,340 plus benefits
  • Department Department of Electrical and Electronic Engineering
  • Category Researcher / Non Clinical Researcher

Job description

Job summary

We have a research associate (postdoc) position to work on spatial audio processing and spatial hearing using methods from machine learning. The aim of the project is to design a method for interactively fitting individualised filters for spatial audio (HRTFs) to users in real-time based on their interactions with a VR/AR environment. We will use meta-learning algorithms to minimise the time required to individualise the filters, using simulated and real interactions with large databases of synthetic and measured filters. The project has potential to become a very widely used tool in academia and industry, as existing methods for recording individualised filters are often expensive, slow, and not widely available for consumers.

The role is initially available for up to 18 months, ideally starting on or soon after 1st January 2022 (although there is flexibility). The role is based in the Neural Reckoning group led by Dan Goodman in the Electrical and Electronic Engineering Department of Imperial College. You will work with other groups at Imperial, as well as with a wider consortium of universities and companies in the SONICOM project (€5.7m EU grant), led by Lorenzo Picinali at Imperial.

Duties and responsibilities

You will:

  • Design and test meta-learning algorithms for spatial hearing in simulations and with human participants.
  • Design simulated environments and players using models of human binaural hearing.
  • Design a VR/AR environment/game and test it on human participants.

You will be supported by other teams in the consortium on virtual reality, acoustics, psychophysics and modelling of the binaural system. You will:

  • Liaise regularly with these teams to use the results of their work and make your results available to them.
  • Present your findings both internally at project meetings and at conferences and workshops, and submit publications to refereed journals.

You will also have the opportunity to assist in supervising undergraduate and graduate research projects, as well as teaching.

Essential requirements

We are looking for applicants with a PhD in spatial audio, audio technologies, acoustics or machine learning, or a related discipline. Ideally, you will have one or more of the following:

  • Experience of applying methods from machine learning (ideally in meta-learning algorithms, although this is not essential).
  • Experience with spatial audio (for example, HRTFs, models of the binaural auditory system).

In addition, you will have:

  • Published high quality papers in machine learning or spatial audio/hearing.
  • Excellent programming skills, especially in Python for machine learning
  • Excellent verbal and written communication skills.
  • Willingness to work as part of a team and to be open-minded and cooperative both internally and with external project partners.

Further information

*Candidates who have not yet been officially awarded their PhD will be appointed as Research Assistant within the salary range £36,045 - £39,183 per annum.

For informal enquiries about the post please contact Dan Goodman at d.goodman@imperial.ac.uk.

Our preferred method of application is online via our website by clicking ‘apply’ below or go to https://www.imperial.ac.uk/job-applicants/ and search using reference number XXX.

Queries regarding the application process should go to Joan O’Brien at j.obrien@imperial.ac.uk.

Further information about the post is available in the job description.

Interviews will take place in the weeks shortly following the closing date of this advert.