Skip to main content

Communication and Listening Effort for Blind and Partially Sighted People

Categories:

Project outline

Blind and partially sighted people are heavily reliant on gathering information via aural means. However, having reduced visual input increases listening effort. You will either investigate the cocktail party effect for people with sight loss or reducing listening effort for screen readers at high word rates. 

Project Partners

This project is based at the University of Salford with funding from the Royal National Institute of Blind People (RNIB)

A University of Salford logo on a light background, featuring a black and white shield with a stylized rose, sun and chain motif on the left, separated by a red vertical line from the text “University of Salford MANCHESTER” on the right, with “Manchester” in red lettering.

Further project information

You will explore one of the following projects:

  1. Cocktail party effect for blind and partially sighted people
    Being able to follow a group conversation in a noisy environment is important in restaurants, pubs and other social situations. When participants have reduced visual cues, this makes turn-taking and picking out the speech-in-noise harder. You will start with researching how conversations work in such situations. You will then explore and develop technology that aids those with reduced sight. Inspiration may come from technologies used by people with hearing loss.
  2. Reducing listening effort for screen readers at high word rates
    Screen readers are an important assistive technology for people with sight loss. Screen readers convert content presented visually into speech. Often high word rates are used to increase the speed at which information is communicated. However, this is mentally tiring.

You will first use biophysical measurement and self-report to quantify and evidence the listening effort problem and identify different user profiles. Then you will explore how to reduce listening effort through (1) optimising the speech synthesis and (2) the use of Large Language Models to optimise information delivery.

Subject Areas

  • Physics
  • Computer Science
  • AI
  • Machine Learning
  • Human Computer Interaction
  • Acoustic Engineering
  • Electronic Engineering
  • Linguistics
  • Mathematics
  • Speech Science
  • Acoustics
  • Psychology

Required qualifications/skills

Essential

  • You should have or expected to again an Honours or Master’s degree in Acoustics, Engineering, Physics, Psychology, Computer Science, Neuroscience or a related field.
  • Skills in programming will be required.

Desirable

  • Some familiarity with machine learning, assistive technology or speech science.

Student

Commencing in October 2026

Supervisors