EURASIP Journal on Audio, Speech, and Music Processing

Table 7 Evaluation over all instruments and all Plackett-Burman designs for the proposed aggregation approach (factors as in Table 5)

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

	a		b		c
Fit	\(R^{2} = 0.13, R_{a}^{2} = -0.09\)		\(R^{2} = 0.65, R_{a}^{2} = 0.56\)		\(R^{2} = 0.61, R_{a}^{2} = 0.51\)
Factors	Estimates	p value	Estimates	p value	Estimates	p value
(Intercept)	0.8448	<2e ⁻¹⁶	0.6815	<2e ⁻¹⁶	0.6945	<2e ⁻¹⁶
Mean interval	−0.0015	0.90	−0.0041	0.76	0.0308	0.17
Onsets accompaniment	−0.0021	0.87	−0.0636	4e ⁻⁰⁵	−0.0448	0.05
Dynamic	−0.0146	0.25	−0.0186	0.17	−0.0019	0.93
Accompaniment	0.0177	0.16	−0.0109	0.41	−0.1313	2e ⁻⁰⁶
Mean pitch	0.0029	0.81	0.0510	6e ⁻⁰⁴	0.0198	0.37
Tone duration	−0.0087	0.49	−0.0348	0.01	−0.0026	0.91
Pitch: mel. - acc.	0.0051	0.68	−0.0224	0.10	−0.0213	0.34

The average F value is the target variable—a: monophonic onset detection, b: predominant onset detection, and c: polyphonic onset detection (italics = significant at 10% level)

Back to article page