Wednesday 7 November 2018, 4.00PM
Speaker(s): Amy Beeston (University of Leeds)
‘Audio-sensing in participatory sound art: Perceptually- informed methods for room-robust sound information retrieval’
In this talk I will discuss the topic of audio-sensing, or machine listening, in the context of participatory sound art. I argue that while many sound artists are intimately concerned with the processes of human listening, relatively few have yet engaged with machine listening in this context. In part, this may be due to two significant difficulties which have not yet been fully investigated: firstly, machine listening software must be tuned to the artistic idea itself (e.g. to detect specific sounding events, or to track the voices of gallery visitors); secondly, methods must be robust to background noise and reverberation in order to deal 'sensibly' with ever-present audio signal distortions (e.g. from other gallery visitors and from unknown room acoustic conditions). Drawing examples from my own and others' work, I hope to facilitate sound artists' engagement with audio-sensing methods by showing how a psychoacoustically-motivated approach to machine listening – using insights from human audition – can help to create more 'reliable' techniques for use in participatory sound works in art gallery settings.
Amy Beeston is a postdoctoral research fellow in Music Psychology at the University of Leeds, and a visiting academic in Computer Science at the University of Sheffield. During her PhD she investigated processes by which human listeners compensate for reverberation in order to adapt to new acoustic environments. Since then she has worked in projects that encapsulate principles of human audition in machine listening software for educational, clinical and industrial applications. She is a co-founding member of SONA and the Yorkshire Sound Women Network which aim to inspire and enable more women and girls to explore sound and music technology.
Location: D/003, Sally Baldwin D Block