Heading hidden


The Artificial Creative Intelligence and Data Science (ACIDS) team is part of the IRCAM laboratory and is hell-bent on trying to model musical creativity by developing innovative artificial intelligence models. Studying creativity and musical improvisation can be seen as the epitome of symbiotic interactions, which can allow to move forward decisively in modeling improvisation in general human behavior. Hence this research aims to provide breakthrough tools for new leaps in musical creativity. Furthermore, having artificial intelligence models that can demonstrate creative behaviors could give rise to a whole new category of generic creative learning systems. The major goal of this team is to produce creative models that can be useful for humans to create inspiring art.


Research projects

We are currently actively developing and collaborating on research on artificial creativity

Learn more about our projects


Open-source models and data

We are committed to open source and open data

Our team produces open-source models, downloadable softwares and open-data sets that are distributed freely for anyone to use.

Fully open source research projects

Accelerating research

All of our research projects are entirely open source. You can find the corresponding code in our GitHub


Open data for innovative creativity problems

We introduced several questions surrounding human creativity and how to model it. We share all our data openly

Finalized software for augmented creativity

Our end goal is to produce usable software for creativity, that we also share for free

Download datasets and codes


Creative outputs

Our projects are oriented towards assisting creativity in music composition

Our research work has been featured in a wide variety of contemporary music production, and several of our software were used for several musical compositions from well-known composers such as Jonathan Harvey, Yan Maresz and Marco Suarez. Recently, our team worked towards the use of generative AI models in order to produce a synthetic opera singer. This work was featured in the opera « La machine des monstres » by Daniele Ghisi. The goal of this research was to be able to hear the learning procedure of an AI, where the model represents a modern version of Shelley’s Frankenstein.