Using Machine Learning to Build Musical Instruments in the Browser

Exploring innovative digital music, art and research which investigates the impact of networking technology on musical creation and performance practice,Network Music Festivalpresents cutting edge musical performances, workshops and discussions.The Network Music Festival took place 2012-2014 at the heart of Birmingham’s (UK) creative community. Responding to the move to online music making during the 2020 global pandemic and with increasing concern in artistic communities about the climate emergency, the festival is returning for aglobalfourth edition, which will take place entirely online15-18th July 2020.

Network Music Festival

Use machine learning to map inputs to outputs and develop a new musical instrument within the MIMIC browser based platform.

Interactive Machine Learning (IML) is a great approach for building interactive systems, including new music instruments and their mappings from input to sound. Building browser based tools is useful because Javascript is a fast growing language with low barriers to entry, additionally, browser based tools require little of the installation and dependencies that plague other data science endeavours.In this workshop, we teach participants to use machine learning to map an input (video, audio, motion…) to an output (MaxiSynth and MaxiSampler) and develop a new musical instrument all within the MIMIC browser based platform. We plan a simple tutorial of machine learning concepts, followed by an introduction to the website and tools themselves, followed by an open workshop where participants develop their own classifiers and mappings under the guidance of workshop leaders.Level: intermediate==========================================================Donations on a Pay What You Can basis. Suggested donation scale: £1 (unwaged)- £20 (those with spare income or institutional support)If for any reason you have trouble making a donation or require a bursary to attend please get in touch with us at networkmusicfestival@gmail.com and we can register you manually. ==========================================================This registration page is for those who intend to be active participants in the workshop. Active participation in the workshops will take place on video conferencing software such as BigBlueButton and Jitsi. We also intend to simultaneously livestream the workshop to our main stage website live.networkmusicfestival.org to facilitate broader passive participation. This live stream will not be publicly archived. If you would like to register as an active participant but DO NOT wish to be included in the livestream please email us to discuss this.==========================================================Louis McCallum is an experienced software developer, researcher, artist and musician. Currently, he is an associate lecturer and researcher at the Embodied AudioVisual Interaction Group, Goldsmiths, University of London. Recent work has involved developing the IML tool Wekinator with Dr Rebecca Fiebrink, and also as a researcher on the AHRC funded MIMIC project (https://mimicproject.com) building web based tools for musicians to use machine learning. Currently teaching a module on machine learning for creative practitioners, he is excellently placed to provide insight into how machine learning can be used as a design tool for musicians, as well as the ways state of the art machine learning techniques can be incorporated into musical practice.Mick Grierson is Research Leader at UAL Creative Computing Institute (CCI). His research explores new approaches to the creation of sounds, images, video and interactions through signal processing, machine learning and information retrieval techniques. He is Principal Investigator on the £1million Machine Learning and Artificial Intelligence project MIMIC, which is a partnership between CCI, Goldsmiths, Durham University, Sussex University, and Google’s Project Magenta. He is also lead academic researcher on the £1million Wellcome Hub project “Heart and Soul at the Hub”, exploring technology assisted research methods with learning disabled and autistic people, disabled arts charity Heart n Soul, RCA and UCL. Hardware and software based on his research has been widely used by world leading production companies, tech start-ups and artists including the BBC, Channel 4, Sigur Ros, Christian Marclay, Martin Creed, Arca and many others. Mick is also the director of the Daphne Oram Collection, and co-founder of the Daphne Oram Trust.