Music research

The Music team at Sony Computer Science Laboratories Paris was set up in 1997 by François Pachet.

The Music Team is concerned with two main areas: the multiple facets of interaction in music, and the challenges of robust music description.

Since its creation, the team developed several pioneering technologies (constraint-based spatialization, intelligent music scheduling using metadata) and award winning systems (MusicSpace, PathBuilder, The Continuator for Interactive Music Improvisation, VirtualBand, Virtuoso).

This work leads to new modes of access to music, interaction with sound, and human interaction. Our systems are expanding towards a new generation of tools, the Flow Machines, whose aim is to abstract "style" from concrete corpora (text, music, etc.), and turn it into a malleable substance that acts as a texture. Applications range from music composition to text or drawing generation. Technically, these systems raise issues in machine-learning (learning style, extracting features), combinatorial optimization (fitting styles to arbitrary constraints) and knowledge representation.

Projects

Flow Machines

Continuator

Virtuoso

VirtualBand

Praise

Miror

Markov Constraints

BackJava

Musaicing

MusicSpace

Feature Generation

Timbre

Playlist

Active Listening

Animal Communication

Publications

Hadjeres, G., Pachet, F. and Nielsen, F. DeepBach: a Steerable Model for Bach Chorales Generation. In Doina Precup and Yee Whye Teh, editor, Proceedings of the 34th International Conference on Machine Learning, Proceedings of Machine Learning Research (vol. 70), pages 1362-1371, International Convention Centre, Sydney, Australia, August 2017 PMLR.
2017
download document
view details
Pachet, F., Papadopoulos, A. and Roy, P. Sampling Variations of Sequences for Structured Music Generation. Proceedings of the 18th International Society for Music Information Retrieval Conference, pages 23-27, Suzhou, China, October 2017 ISMIR.
2017
download document
view details
Sakellariou, J., Tria, F., Loreto, V. and Pachet, F. Maximum entropy models capture melodic styles. Scientific Reports, 7(9172), August 2017
2017
download document
view details
Roy, P., Papadopoulos, A. and Pachet, F. Sampling Variations of Lead Sheets. arXiv:1703.00760, March 2017 16 pages, 11 figures
2017
download document
view details
Marchini, M., Pachet, F. and Carré, B. Rethinking Reflexive Looper for structured pop music. To appear in The International Conference on New Interfaces for Musical Expression, 2017
2017
download document
view details
Pachet, F. Roy, P. Do jazz musicians really interact?. In Micheline Lesaffre, Pieter-­Jan Maes and Marc Leman, eds., editor, Routledge Companion to Embodied Music Interaction, Routledge. 2017
2017
view details
Hadjeres, G and Pachet, F. DeepBach: a Steerable Model for Bach chorales generation. arXiv:1612.01010, December 2016 https://arxiv.org/abs/1612.01010
2016
download document
view details
Gorlow, S., Ramona, M. and Pachet F. Decision-Based Transcription of Jazz Guitar Solos Using a Harmonic Bident Analysis Filter Bank & Spectral Distribution Weighting. arXiv:1611.06505, November 2016 https://arxiv.org/abs/1611.06505 Audio Files: https://doi.org/10.5281/zenodo.167567
2016
download document
view details
Gorlow, S., Ramona, M. and Pachet F. SISO and SIMO Accompaniment Cancellation for Live Solo Recordings Based on Short-Time ERB-Band Wiener Filtering & Spectral Subtraction. arXiv:1611.08905, November 2016 https://arxiv.org/abs/1611.08905
2016
download document
view details
Hadjeres, G., Sakellariou, J. and Pachet F. Style Imitation and Chord Invention in Polyphonic Music with Exponential Families. arXiv:1609.05152, September 2016 http://arxiv.org/abs/1609.05152
2016
download document
view details