Skip to main content
School of Electronic Engineering and Computer Science
  1. School of Electronic Engineering and Computer Science
  2. PhD
  3. PhD studentships
  4. Deep neural network models of music perception, from sound to pleasure

Deep neural network models of music perception, from sound to pleasure

Project Title: Deep neural network models of music perception, from sound to pleasure 

Supervisor: Dr Marcus Pearce (https://www.marcus-pearce.com)  

Project Description: Why does the same piece of music thrill one listener and leave another cold? This PhD will develop deep neural network models of how people perceive and enjoy music, and use them to explain why listeners of different ages and cultures experience music differently. 

The project builds on our laboratory’s research programme showing that much of musical experience can be simulated by models that learn to predict what comes next in music. Trained on the music a listener has heard, these models reproduce human expectations, the perception of thematic structure, and judgements of complexity and pleasure (see publications). The PhD will take this approach into deep learning, with three aims: 

  • Learning from sound. Develop models that learn to predict directly from audio, closer to what listeners actually hear than current approaches that process symbolic representations of music. 
  • From perception to pleasure. Build hierarchical models that connect low-level auditory processing, through musical structure, to the experience of pleasure. 
  • Interpretation of learned representations. Examine what the embeddings in trained models encode about musical structure and how they relate to perceptual representations. 

Models will be tested and refined by comparing their responses with behavioural and neural data from human listeners. The work will advance the science of musical experience and lay the foundations for systems that recommend music tailored to individuals, in everyday life and in education and healthcare. 

The successful candidate will have an opportunity to shape the project’s direction. Possible directions include: 

  • Differences between listeners: cross-cultural differences, individual differences in perception and pleasure, and developmental change from childhood to old age. 
  • Brain and mind: simulating neural processing of musical structure, and comparisons between music and language. 
  • Musical culture: how musical styles evolve in simulated populations of listeners and composers. 

The studentship will be based in the Computational Auditory Perception Lab at Queen Mary University of London. The lab is part of the Centre for Digital Music, one of the world’s leading research groups in music and audio technology. It also belongs to two of the university’s interdisciplinary research centres: the Centre for Multimodal AI and the Centre for Human-Centred Computing. The student will join a community of researchers working across music perception, cognitive science, and artificial intelligence. Flexible working arrangements, including remote working, can be discussed.  

Primary Objective: Develop and validate deep neural network models that simulate how listeners perceive and enjoy music, from the audio signal to the experience of pleasure, and that account for how perception differs between listeners with different musical experience. 

Secondary Objectives:  

  • Build predictive models that learn musical structure directly from audio, and test whether they match or outperform existing symbolic models in predicting listeners’ expectations. 
  • Link levels of processing in a hierarchical model: auditory features, then musical structure, then judgements of complexity and pleasure. Test each level against behavioural ratings from human listeners. 
  • Interpret what trained models learn by relating their internal representations to musical structure and to neural responses recorded while people listen to the same music. 
  • Simulate differences between listeners by training models on different musical exposure (cultures, ages or individual listening histories), and test whether the trained models reproduce the corresponding differences in human perception and pleasure. 
  • Analyse and model trajectories of cultural change in music using artificial simulations of societies of listeners.   

Prerequisites: 

  • A Master’s degree at Distinction level or equivalent (completed, or expected to be completed before the PhD starts) in computer science, artificial intelligence, psychology, cognitive science, neuroscience, computational neuroscience, computational linguistics or a related discipline, including maths, physics and engineering. We also welcome applicants with a background in music who have programming or quantitative experience.  
  • A deep interest in how music, the mind and the brain can be understood through computational modelling. 
  • Strong programming skills and familiarity with both conventional machine learning and deep learning.  

Funding: 

The PhD studentship is funded by an EPSRC Doctoral Landscape Award and is open to applicants with Home and International fee status. The number of students with international fee status who can be recruited is capped under EPSRC terms and conditions, so competition for international places is particularly strong. Awards are tenable for up to 3.5 years and cover tuition fees and a maintenance stipend at the applicable UKRI rate, including the London allowance. The indicative stipend for 2026/27 is £23,805 per annum for full-time study, with the 2027/28 rate to be confirmed.  

How to apply: 

Queen Mary University of London welcomes applications from candidates excited by the intersection of AI, music and the human mind. Applicants are encouraged to discuss their research interests with the prospective supervisor before applying and should follow the instructions at: 
https://www.qmul.ac.uk/eecs/phd/how-to-apply/  

The application should include: 

  • CV, maximum two pages. 
  • Cover letter, maximum 4,500 characters, explaining your suitability for the project and indicating your anticipated Home or International fee status. 
  • Statement of Research Interests, maximum 500 words, explaining your interest in the proposed research area and your relevant experience. 
  • Two references. 
  • Evidence of English language proficiency, where required. 
  • Academic transcripts, degree certificates and any other required supporting documents. 

Fee status and studentship eligibility will be assessed by Queen Mary in accordance with the applicable requirements. 

 Application Deadline: 16:00 GMT, 27 January 2027.  

For specific enquiries, contact Dr Marcus Pearce at marcus.pearce@qmul.ac.uk including a CV and a few sentences on which research directions interest you. For general enquiries, contact eecs-phd-enquiries@qmul.ac.uk, using the subject “EECS 2027 PhD studentships enquiry”. 

Back to top