COCHONUT: RECOGNIZING COMPLEX CHORDS FROM MIDI GUITAR SEQUENCES

摘要: Chord recognition from symbolic data is a complex task, due to its strong context dependency and the large number of possible combinations intervals which chords are made of, specially when dealing with dissonances, such as 7 ths , 9 13 suspended chords. None current approaches deal complexity. Most them consider only simple chord patterns, in best cases, including sevenths. In addition, considering captured MIDI guitar, we need non quantized noisy data, increases difficulty task. The no automatic technique reduce noise. This paper proposes new approach recognize chords, guitar called COCHONUT (Complex Chords Nutting). system uses contextual harmonic information solve ambiguous integrated other techniques, decision theory, optimization, pattern matching rule-based recognition. results encouraging provide indications that use information, can actually improve currently found literature.