Data Study Group Final Report: IEDE Acoustics Group, University College London

Deep Learning Techniques for noise annoyance detection (DeLTA)

Abstract

Noise annoyance is often reported as one of the main adverse effects of noise exposure on human health. Chronic high noise annoyance impacts 22 million people in Europe alone, with a broad range of public health outcomes. This Data Study Group applied sound source identification and deep learning methods on a set of urban recordings to create a model which can predict the resulting annoyance rating. The research challenge was to investigate to what degree the inclusion of sound source information can inform the optimal modelling strategy for automatically predicting noise annoyance.

Citation information

DOI: 10.5281/zenodo.10090651

Additional information

Contributers:

  • Emmeline Brown is a fourth-year PhD student affiliated with the Centre for Computational Medicine, UCL
  • Ratneel Deo (Facilitator) Ratneel is a PhD scholar at the University of Sydney. 
  • Yuanbo Hou is a PhD student at Ghent University 
  • Jasper Kirton-Wingate is a 1st year PhD student at Edinburgh Napier University
  • Jinhua Liang is a 2nd-year Ph.D. student at the Centre for Digital Music, Queen Mary University of London. 
  • Alisa Sheinkman is a third-year PhD candidate studying at the School of Math at the University of Edinburgh. 
  • Christopher Soelistyo recently finished a PhD at University College London, and will soon begin a postdoctoral position at the Alan Turing Institute.
  • Hari Sood (Facilitator) is a Research Application Manager at the Alan Turing Institute
  • Arin Wongprommoon is a fourth-year PhD student at the Centre for Engineering Biology, University of Edinburgh. 
  • Kaiyue Xing works as Laboratory Manager at School of Education, Communication and Language Science, Newcastle University. 
  • Wingyan Yip is a Business Intelligence Specialist at Soldo Ltd.