challenging-ai
Can we do the same with less - AI in 64 Kb
Philippe Esling (FR)
This hands-on workshop for the AIxMusic Hackathon by Philippe Esling (FR) introduces techniques for lightweight AI, demonstration of embedded technologies and a 64 Kb competition for an AIxMusic hackathon project challenging the current limits of AI and inspired by the Demoscene and the 64Kb competitions.
AI Creative Agents: The Man I Love
Hervé Sellin (FR), Remi Fox (FR)
The Musical Representations team in the STMS laboratory at IRCAM is working on "cyber-human" systems. Just as cyber-physical systems create a continuity between the digital logic of computers and the physical world by capturing, modeling, and modifying it, they establish a creative continuity between machines and musicians using mathematical modeling and artificial intelligence.
Little Etudes for Piano
Elaine Chew (US/GB)
Little Etudes is a series of short piano pieces based on cardiac electrical anomalies. Aberrant electrical activity in the heart can cause the heart to beat irregularly. Abnormal heart rhythms form interesting musical patterns, which raises the question of whether all natural-sounding musical rhythms have a physiological basis.
Three Ladies Project
Georges Bloch (FR)
Billie Holiday, Édith Piaf and Elisabeth Schwarzkopf are three great ladies of music, born in 1915. Could we make them sing together? This was the subject of a performance conceived for the music festival Le pietre che cantano in L'Aquila (Italy) in 2015. Using DYCI2Lib software, the three ladies finally sing The Man I Love together. ImproteK, the earlier version of DYCI2Lib, already made it possible to intelligently link the harmony of the base material used (called "memories") to the harmonic progression of the reference song (called "scenario").
Talk: Frontiers of Music and Artificial Intelligence
IRCAM (FR)
Artificial intelligence plays an increasingly prominent role in music composition, performance, and production. How does this benefit or undermine musicians? Where are such developments taking us? Two young protagonists at the leading edge of AIxMusic will debate the limits and future directions of AIxMusic, tackling topics ranging from classic questions of AI takeover to therapeutic and gaming applications to discussions of lightweight deep AI, of uncertainty and unknown models and dimensions.
Talk: Musical Generativities (Facilitated by H.Vinet)
IRCAM (FR)
Techniques for the automatic generation of music – which have long been focused on systems ruling the score – are now being deployed at all levels of sound representation: signal, gesture, symbol, form. Sound synthesis using deep networks constitutes a radical break with conventional modelling approaches. How do composers handle these emerging possibilities?
