Speaker: Andrew Mitchell, University College London
Wed 14th May 12:00 – 13:00 UK time. This seminar will be held in person in Peel 102 and you can join us online via Microsoft Teams.
Abstract
This seminar explores the development and application of open source software tools for soundscape analysis and assessment. Drawing from my experience creating Python packages for soundscape research, I will demonstrate how open source approaches enhance reproducibility, accessibility, and collaboration in scientific research. The presentation will showcase Soundscapy, a specialized Python library designed to analyze and visualize soundscape assessments according to ISO standards while addressing common interpretability challenges. I’ll demonstrate how Soundscapy implements innovative visualization approaches for representing collective perceptions through probabilistic distributions within the circumplex model, alongside powerful tools for psychoacoustic and acoustic analysis of binaural recordings. The seminar will also touch on complementary packages I’ve developed and the experiences and challenges of maintaining open source software. Throughout the talk, I’ll highlight how these tools embody the principles of open research that extend beyond mere code sharing to encompass reproducible methods, accessible data, and collaborative development practices. By diving beneath the “open source iceberg,” attendees will gain insights into how these specialized tools can streamline their own research workflows while contributing to more transparent and rigorous scientific practices in acoustics and environmental analysis.
Biography
Dr. Andrew Mitchell is a Lecturer in Artificial Intelligence and Machine Learning for Sustainable Construction at University College London. His research interests include general uses of AI in the construction industry, the application of machine learning methods to environmental analysis, and soundscape engineering. His current ongoing projects include the Soundscape Indices (SSID) project and Soundscape Attributes Translation Project (SATP), as well as several open source Python packages, including Soundscapy, Circumplex, and CitySeg.