EURASIP Journal on Audio, Speech, and Music Processing

Table 6 Results (mean F measure) for onset detection with and without an auditory model (AM)

From: A computational study of auditory models in music recognition tasks for normal-hearing and hearing-impaired listeners

Design	All	Melody	Monoph.
W/o AM
Cello	0.65	0.57	0.80
Clarinet	0.79	0.72	0.80
Trumpet	0.87	0.84	0.97
Mean	0.77	0.71	0.86
AM, best ch.
Cello	0.44	0.37	0.68
Clarinet	0.65	0.61	0.80
Trumpet	0.70	0.79	0.99
Mean	0.60	0.59	0.82
AM, aggr.
Cello	0.53	0.46	0.79
Clarinet	0.71	0.72	0.76
Trumpet	0.85	0.87	0.98
Mean	0.69	0.68	0.84

Back to article page