Ceci est une ancienne révision du document !


ACE Analyzer

We developed an ACE Analyzer including two analysis modules to discover musical relationships between target chords and chords predicted by Automatic Chord Extraction systems.

The first module detects the errors corresponding to hierarchical relationships or usual chord substitutions rules: using a chord in place of another in a chord progression (usually substituted chords have two pitches in common with the triad that they are replacing).

The second module focuses on harmonic degrees. First, by using annotations of key in addition to that of chords, this module determines the harmonic degrees of the predicted chord and of the target chord. Then, it counts the substitutions of harmonic degrees when it is possible.

Download Python ACE Analyzer (coming soon)

More information: refer to Carsault, Nika, Esling, “Using musical relationships between chord labels in Automatic Chord Extraction tasks”, in Proceedings of the 19th International Society of Music Information Retrieval Conference, Paris, France, 2018 (please mention this paper when using the library).

 


dyci2/ace_analyzer.1528898158.txt.gz · Dernière modification: 2018/06/13 15:55 par Jérôme Nika