About

nume­di­art is a long-term research pro­gramme cen­tered on Dig­i­tal Media Arts, funded by Région Wal­lonne, Bel­gium (grant N°716631). Its main goal is to fos­ter the devel­op­ment of new media tech­nolo­gies through dig­i­tal per­for­mances and instal­la­tions, in con­nec­tion with local com­pa­nies and artists.

It is orga­nized around three major R&D themes: HyFORGE – hyper­me­dia nav­i­ga­tion, COMEDIA – body and media, COPI – dig­i­tal instru­ment mak­ing. It is per­formed as a series of short (3-months) projects, typ­i­cally 3 or 4 of them in par­al­lel, which are con­cluded by a 1-week “hands on” work­shop. Par­tic­i­pa­tion to projects is open, and can be done remotely, but phys­i­cal par­tic­i­pa­tion to the work­shop is manda­tory. No fund­ing is pro­vided, but no fee is asked for either (check Calls for Par­tic­i­pa­tion for fur­ther details).

nume­di­art is the result of col­lab­o­ra­tion between Polytech.Mons (Infor­ma­tion Tech­nol­ogy R&D Depart­ment) and UCL (TELE Lab), with a cen­ter of grav­ity in Mons, the cul­tural cap­i­tal of Wal­lo­nia. It also ben­e­fits from the exper­tise of the Mul­ti­tel research cen­ter on mul­ti­me­dia and telecom­mu­ni­ca­tions. As such, it is the R&D com­po­nent of MONS2015, a broader effort towards mak­ing Mons the cul­tural cap­i­tal of Europe in 2015.

http://www.dailymotion.com/videox8v3iv

The kick-off slideshow can be down­loaded here.

R&D Axes

HyFORGE — Hyper­me­dia Nav­i­ga­tion (read more…)

Infor­ma­tion index­ing and retrieval rely clas­si­cally on con­strained lan­guages to auto­mat­i­cally describe con­tents and allow for­mu­lat­ing queries, respec­tively. This approach becomes hardly applic­a­ble for mul­ti­me­dia con­tents such as music or video because of the dis­par­ity between com­putable low-level descrip­tors and desired high-level seman­tics — the so-called seman­tic gap. Alter­na­tively, HyFORGE inves­ti­gates human-in-the-loop approaches and inno­v­a­tive tools for struc­tur­ing and search­ing mul­ti­me­dia con­tents. Along with audio and image pro­cess­ing, HyFORGE builds up on self-organizing mod­els to derive enhanced views of mul­ti­me­dia col­lec­tions and pro­vide users with effi­cient brows­ing interfaces.

http://www.dailymotion.com/videox8v2p6

COMEDIA — Body & Medias (read more…)

Come­dia is named from a French con­trac­tion between body and media or stage direc­tor and media, which nicely sums up the main objec­tif of this axis: giv­ing to bod­ies the means to be their own artis­tic direc­tor! Hence based on posi­tion on stage or choreg­ra­phy between mul­ti­ple artists for the inter-relationship and ges­tures or voice for the intra-relationship, Come­dia aims at cre­at­ing inter­ac­tiv­ity between per­form­ing artists and the mul­ti­me­dia con­text around. Event descrip­tion, low-level fea­ture analy­sis, pat­tern recog­ni­tion, het­ero­ge­neous sen­sor fusion, robust­ness against light­ing and real-time are our key­words in 1D, 2D and 3D sig­nal pro­cess­ing to reach these goals.

http://www.dailymotion.com/videox8e31l
http://www.dailymotion.com/videox94pbe

COPI — Dig­i­tal Instru­ment Design (read more…)

COPI aims at devel­op­ing a software/hardware tool­box for cre­at­ing inno­v­a­tive dig­i­tal musi­cal instru­ments, from scratch or by aug­ment­ing exist­ing instru­ments with new inter­ac­tive chan­nels. The main chal­lenges for this R&D axis are to pro­duce expres­sive instru­ments which main­tain a close, embod­ied rela­tion­ship with the musi­cian. Our approach is to pro­duce new sound design archi­tec­tures using a large data­base of pre-recorded sig­nals while main­tain­ing real-time con­trol of the design process. Our sci­en­tific work there­fore implies three main axes: the devel­op­ment of expres­sive pro­duc­tion mod­els (audio sig­nal pro­cess­ing), fol­lowed by the design of ges­tural con­trol sys­tems for their syn­the­sis para­me­ters, cou­pled with sta­tis­ti­cal mod­el­ing of this dynamic control.

http://www.dailymotion.com/videox8v38f
http://www.dailymotion.com/videox8v3bb