School of Electronic Engineering and Computer Science

Dr Dan Stowell



Telephone: +44 20 7882 3066
Room Number: Engineering, Eng 400


Deep Learning for Audio and Music (Postgraduate)

This module, for those who have some prior knowledge of machine learning, focusses on deep learning methods and how they can be used to address many tasks in audio and music. The theory of modern deep neural networks (DNNs) is covered, including training of common DNN types as well as modifying DNNs for new purposes. Various tasks in analysis/generation of audio and music are studied directly to inspire the content, using raw audio and/or symbolic representations. Background in machine learning is essential, and some background in digital signal processing is highly recommended. Music knowledge would be desirable but is not a requirement.


Research Interests:

I am a researcher interested in machine learning applied to sound. This spans the range from automatic recognition of sounds such as birdsong, to creative applications such as human-machine on-stage performance.

For publications and more info see