Research & Innovation

Designing the tools that invent new sounds and languages

(research laboratories, creative studios, technology brands, institutions seeking a sound signature generated in real time rather than a static audio file)

Fred Voisin has long taken part in institutional research projects spanning ethnography to neuroscience and sound design. He began working in computer music in the Central African Republic and Indonesia, studying traditional music for the CNRS in the early 1990s, as cognitive science was taking off. Building on that experience, in 1995 Fred developed, in collaboration with composer Jacopo Baboni-Schilingi, tools for the analysis and generation of forms for the OpenMusic environment — a contribution praised by a community of young composers.

And, well before the year 2000, before generative AI became a mainstream topic, Fred started the Neuromuse project, a "system of neuromimetic networks for musical creation." It was a pioneering project in generative AI that still allows — today — the instantiation of neuromimetic agents able to communicate over MIDI or TCP/IP. In Lisp, and in the Max and Puredata multimedia environments, with further adaptations in Python, Java, C++ and ARM... It led to proofs of concept in the form of experimental-music concert-performances, co-produced with various research centres such as Ircam, CIRM, Art Zoyd, but also Telecom Valley, Inria, France Telecom, Eurecom, Infineon, National Taiwan University, and others — see the Neuromuse project on GitHub.

Musical research, understood across multiple horizons, makes it possible to design tools that are genuinely bespoke, tailored to research and innovation.

• Development of bespoke generative systems (neuromimetic networks, adaptive synthesis engines)
• R&D in artificial intelligence applied to sound creation and identity
• Prototyping and advice for experimental or applied-research projects
• Support for laboratories, studios or technical teams wishing to integrate a generative sound component

Productions

Tools & software

• SpeaK for Max — Max patch for SpeaK, a collaborative platform for sound lexicons, developed with Patrick Susini (graphic design and ergonomics), Ircam - Perception and Sound Design team (2022) — speak.ircam.fr
• LabanOnLisp (LOL) — experimental environment for choreographic composition in Lisp (1999-2012), developed notably with Myriam Gourfink — documentation
• Neuromuse — neuromimetic network system for musical creation (code repository) — github.com/FredVoisin/neuromuse
• Morphologie — library for OpenMusic (Ircam), inspired by ethnomusicology and René Thom — github.com/openmusic-project/Morphologie, project page

Gesture-sound interaction (LEAD, CNRS - Université de Bourgogne)

• Mimer c'est jouer ! — musical exploration prototype based on miming instrument playing, gesture recognition through a neuromimetic network (MoCap, MLP TDNN), with Emmanuel Bigand, supported by Cochlear (2017) — PDF
• All Together — demonstration of musical sonification of movement through wireless sensors, playing together as a group, with Emmanuel Bigand & Philippe Lalitte, Fête de la Musique, CNRS Gérard-Mégie campus, Paris (2017) — cnrs.fr

Neuromimetic works & installations

• Last Manoeuvres in the Dark — with Fabien Giraud, Raphaël Siboni & Robin Meier, Palais de Tokyo, Paris (2008) — robinmeier.net
• Symphonie des machines — installation-concert for neuromimetic agents, Festival de la 4ème Dimension, Sophia Antipolis (2006), supported by the Ville de Paris, SNCF, Sacem, Art Public Contemporain and CIRM — robinmeier.net

Publications

• « Vocal Imitations of Non-Vocal Sounds » (with Guillaume Lemaitre, Olivier Houix, Nicolas Misdariis, Patrick Susini), PLOS ONE, 11(12), e0168167 (2016) — DOI
• « Playing Live Music Using Artificial Neural Networks » — invited talk, NeuroArts Festival, University of Plymouth (11 February 2011) — archive.org
• « On Analytical vs. Schizophrenic Procedures for Computing Music » (with Robin Meier), Contemporary Music Review, 28(2), pp. 205-219 (2009) — DOI
• « Ethnomusicologie et technologie numérique » — talk at the colloquium « Un passé plein d'avenir : à la rencontre du patrimoine musical » (Université de Saint-Quentin-en-Yvelines / UNESCO), Palais de la Découverte, Paris (4 April 2008)
• « Playing Integrated Music Knowledges with Artificial Neural Networks » (with Robin Meier), Journées d'Informatique Musicale / Sound and Music Computing, Paris (2004) — HAL