The Cuidado music browser: an end-to-end electronic music distribution system.
Conference Proceeding: The SemanticHIFI project: content-based management and manipulation of musical recordings[show abstract] [hide abstract]
ABSTRACT: The SemanticHIFI project aims at designing and prototyping tomorrow's Hi-fi systems, which will provide music lovers with innovative functions of management and manipulation of musical contents. The limitations of current equipments are mainly related to those of the music distribution media (album-based audio recordings in stereo format), with poor control features and interfaces (album/ track selection, play, stop, volume, etc.). Enabling the manipulation of richer media and related metadata (either distributed with the audio recordings or computed by the user using dedicated indexing tools) opens a wide range of new functionalities: personal indexing and classification of music titles, content-based browsing in personal catalogues, browsing within titles with automatic segmentation and de-mixing tools, 3D audio rendering and assisted mixing features, etc. Moreover, the manipulation of interactive music contents will be made accessible to music consumers, through dedicated performing, and authoring tools. They will then have the possibility of publishing and sharing their personal work with others, through a dedicated peer-to-peer sharing middleware specifically designed for preserving the rights of the used digital media. All these features are the result of the various R&D tasks and experiments performed as part of the project and represent the state-of-the art in various research fields : digital audio signal processing, music information retrieval, man-machine interfaces and peer-to-peer networks. Moreover, the project includes an integration phase, which aims at producing full-featured applications prototypes, designed to fit identified market needs, through technical choices compatible with these markets. This article proposes an overview of the project, by presenting its background and objectives, the main scientific issues and breakthroughs it addresses in relation to the description and extraction of musical information, the main applications featur- - es it aims at developing and choices made for their integration into application prototypes compliant with market needs.Integration of Knowledge, Semantics and Digital Media Technology, 2005. EWIMT 2005. The 2nd European Workshop on the (Ref. No. 2005/11099);
catalogues.? First,? we? are? interested? in? human-centred?
and? known? to? many? listeners.? Second,? we? consider?
“popular? browsing”? of? music,? i.e.? making? music?
sharing? of? musical? tastes? and? information? within?
descriptor? extraction? from? the? music? signal,? or? data?
music? retrieval? methods? such? as? automatic? sequence?
generation,? and? user? interface? issues.? This? paper?
the? results? obtained,? in? the? current? state? of? the? IST?
First,? purely? editorial? systems? propose? systematic?
track? listings? (CDDB,? Musicbrainz),? information? on?
(CDDB,? Musicbrainz).? These? systems? provide? useful?
two? basic? techniques:? 1)? an? audio? fingerprinting?
technology? able? to? recognize? music? titles? on? personal?
songs,? which? is? incremented? automatically,? and? in? a?
from? user? ratings? and? does? not? perform? any? acoustic?
systems,? metadata? repositories,? e.g.? Wold? et? al.? 1996)?
ingredients? of? the? music-to-listener? chain,? for? a? fully-
project? covers? the? areas? of? 1)? editorial? metadata,? 2)?
The? next? sections? describe? the? most? important? results?
Consensual? information? or? facts? about? music?
The? first? category? is? common? to? already? existing?
instance:? artist? and? songs? name,? albums? and? tracks?
The? second? category? is? more? problematic.? Content?
tool? allows? editing? and? adding? artists? and/or? songs?
Editorial? metadata? are? associated? distinctly? with? music?
identifiers? for? many? users:? Yesterday? is? by? “The?
Beatles”,? and? “The? 5th? symphony”? is? by? Beethoven.?
“Stabat? Matter”? by? Pergolese? is? not? the? one? by?
they? are? most? commonly? used? to? identify?music?titles.?
important:? not? only? for? interpreters,? but? also? for?
conductors? (for? orchestral? pieces).? In? non-Classical?
Existing? repositories? of? editorial? information? do? not?
duets? (Paul? McCartney? &? Michael? Jacskon).? To? each?
instrumentist,? etc.),? country? of? origin,? language? (for?
(for? instrumentists).? Other? information? concern? the?
relation? MHE? entertain? with? each? other.? For? instance,?
Phil? Collins? a? memberOf? the? group? Genesis.? The?
Editorial? MHE? database? may? be? seen? more? as? a?
as? title? name? or? keywords,? as? well? as? less? obvious?
are? as? badly? ill-defined.? Our? studies? on? existing?
taxonomies? of? genres? have? shown? that? there? is? no?
errors,? we? ended? up? with? a? simple? two-level? genre?
songs? either? in? a? generic? way? (Classical,? Jazz),? more?
simpler? taxonomies? may? also? produce? frustration,? as?
be? entered? by? users? to? further? refine? their? own?
yet? flexible? approach? has? the?advantage?of?uniformity:?
artists? and? songs? are? classified? in? the? same? taxonomy,?
The? Beatles? is? classified? in? “Pop? /? Brit”,? but? Beatles?
The? main? type? of? metadata? that? the? MB? proposes? for?
i.e.? information? extracted? from? the? audio? signal.? The?
Mpeg7? standard? aims? at? providing? a? format? for?
extracted? automatically? in? a? systematic? way.? Typical?
signal? characteristics? such? as? Means? and? Variance? of?
high-level? perceptual? categories,? Mpeg7? is? strictly?
concerned? with? the? format? for? representing? this?
We? have? conducted? in? the? project? several? studies?
focusing? on? particular? dimensions? of? music? that? are?
extract? the? time? series? of? percussive? sounds? in? music?
.? However,? many? things? remain? to? be? done? in?the?
in? how? to? extract? rhythm,? but? how? to? exploit? the?
with? words,? and? even? less? to? produce? rhythm? (our?
pertaining? to? popular? music? access,? the? perceptual?
hard? rock,? dance? music),? or? relaxing? and? calm? (e.g.? a?
We? have? studied? the? correlation? of? experimental?
measures? (user? tests)? with? a? variety? of?signal? features,?
combinations? (using? discrimination? analysis)? and? their?
possible? compositions? with? signal? operators? (filters,?
)))((var( 10logx diff
quite? obvious? that? music? taste?is? often? correlated? with?
sound? of? Chick? Corea? playing? on? an? electric? piano),?
others? to? musical? configurations? (e.g.? the? sound? of? a?
We? model? the? global? “sound”? of? a? music? title? as? a?
distribution? in? the? space? of? mel? cepstrum? coefficients?
(MFCC).? MFCCs? provide? a? compact? representation?of?
the? signal’s? spectral? envelopes,? which? are? a? good?
correlate? of? the? timbre.? By? comparing? timbre?
–? Yesterday”,? and? the? dark-gray? GMM? is? the? timbre?
model? of? the? song? “Joao? Gilberto? –? Besame? Mucho”.?
This? two? songs? have? a? very? similar? “sound”? (acoustic?
male? voice),? and? indeed? we? see? that? their? mfcc?
3.2.?EDS:? A? General? Framework? for? Extracting?
These? various? studies? in? descriptor? extraction? from?
acoustic? signals? have? shown? that? the? design? of? an?
efficient? acoustic? extractor? is? a? very? heuristic? process,?
which? requires? sophisticated? knowledge? of? signal?
processing,? intuitions,? and? experience.? Indeed,? most?
approaches? in? feature? extraction? as? published? in? the?
literature? consist? in? using? statistical? analysis? tools? to?
in? this? category.? However,? these? approaches? are? not?
nature? of? the? palette? of? LLD,? which? usually? do? not?
capture? the? relevant,? often? intricate? and? hidden?
virtually? infinite? number? of? extractors? of? musical?
the? sound? (“some? saturated? guitar? with? a? little? bit? of?
chorus”),? while? another? simply? wants? to? find? “funky”?
These? experiments? have? given? rise? to? a? systematic?
approach? to? feature? extraction,? embodied? in? the? EDS?
idea? of? EDS? is? to? automate? –? in? part? or? totally?–? the?
much? in? the? same? way? than? experts? do:? by? inventing?
functions,? computing? them? on? test? databases,? and?
To? reach? this? goal? EDS? uses? a? genetic? programming?
engine,? augmented? with? fine? grained? typing? system,?
which? allows? to? characterize? precisely? the? inputs? and?
outputs? of? functions.? EDS? uses? also? rewriting?rules?to?
simplify? complex? signal? processing? functions? (see? the?
expert? knowledge? to? guide? its? search,? in? the? form? of?
Typical? heuristics? include? “do? not? try? functions? which?
contain? too? many? repetition? of? the? same? operator”,? or?
times”,? or? also? “spectral? coefficients? are? particularly?
hand),? discrimination? between? songs? and? instrumental,?
harmonic? complexity,? etc.? The? ambitious? goal? of?EDS?
makes? it? a? project? in? itself,? as? it? aims? at? capturing?
think? that? the? contribution? to? the? MIR? community? is?
of? music? information? retrieval,? and? the? expectation? to?
again,? similarity? is? ill-defined,? and? it? can? be? of? many?
different? sorts.? For? instance,? one? may? consider? all? the?
two? titles? may? be? considered? similar? by? a? user? or? a?
community? of? users? for? no? objective? reason,? simply?
similarity? measures? from? the? metadata? obtained? and?
described? above,? either? editorial? or? acoustic.? Most?
is? based? on? the? global? “timbre”? of? the? songs.? The?
distance? analysis? is? based? on? Gaussian? models? of?
sampled? and? then? the? likelihood? of? the? samples? is?
computed? given? the? other? model.? Figure? 5? shows? a?
Browser.? Here,? the? user? has? select? a? jazz? piano? song?
(“Ahmad? Jamal-? L’instant? de? Vérité”),? and? asked? the?
lists? contains? songs?of? many? genres,?which?all?contain?
“Variety”? song? (William? Sheller,? a? French? singer? and?
occurrence? analysis? is? based? on? a? simple? idea:? if? two?
and? spoken? text? has? been? used? to? extract? clusters? of?
semantically? related? words.? Similarity? measurements?
to? be? cognitively? plausible? .? We? have? identified?
Its? goal? is? to? gather? as? many? web? pages? as? possible,?
crawled? web? page? is? given? a? score? according? to? the?
hard? drive? can? contain.? Therefore,? users? can? create?
database? on? specific? topics? or? according? to? specific?
tastes.? For? example,? if? you? interested? in? “intelligent?
can? start? crawling? using? the? first? answers? provided? by?
Google? as? well? as? specific? keywords? you? entered? like?
you? construct? an? “intelligent-techno”? oriented?database?
The? second? part? of? this? software? is? devoted? to? the?
distance? computation.? The? various? formula? used? here?
of? words? in? the? same? page,? taking? into? account? the?
the? compatibility? with? the? Music? browser? users? can?
import? any? data? coming? from? Cuidado? tables.? The?
We? have? covered? so? far? the? core? technologies? for?
producing? content? descriptions? of? music? titles.? A? key?
top? boxes??? PDAs? ?? telephones??? Hard-disk? Hi-FI??).?
Many? user? interfaces? have? been? proposed? for? music?
access? systems,? from? straightforward? feature-based?