Table des matières

Resources

Sound examples, interactive listening tests

WP1 Informed multichannel separation: interference reduction, ICASSP 2015 examples
Separation examples, Remixing examples

WP1 Interactive listening tests
Multichannel source separation with deep neural networks

WP2 Learning of multidimensional and multi-level structures :

Multimedia: performances, experiments, qualitative interviews

WP4 videos of interactions with expert musicians
Public performances with DICY2 prototypes (ImproteK, WP3)
Listening sessions, interviews, qualitative validation of DYCI2 prototypes (ImproteK, WP3) by professional musicians

Section IV in Jérôme Nika's Phd thesis fully describes the sessions of musical experiments and qualitative validations of the DYCI2 prototypes (ImproteK, WP3). This analysis is illustrated with sound and video documents.

Prototypes

Repository of the different prototypes (:!: work in (constant) progress :!:):
DYCI2 library: https://forge.ircam.fr/p/Dyci2_library
ImproteK : https://forge.ircam.fr/p/ImproteK
Somax : https://forge.ircam.fr/p/Somax

Data base

WP2 Charlie Parker's Omnibook Data, encoded by K. Déguernel in Midi, MusicXml, MuseScore and PDF

Million song dataset

Lakh midi dataset (matched and aligned to entries in the Million Song Dataset)

More information...

An update on the Somax project, Laurent Bonnasse-Gahot (pdf).

Guiding human-computer music improvisation with temporal scenarios: summary (french)