Auditory Sparse Representation for Robust Speaker Recognition Based on Tensor Structure

EURASIP Journal on Audio, Speech, and Music Processing

Table 2 Identification accuracy in four noisy conditions (white, pink, factory, and f16) for Grid dataset.

(%)	SNR	ANTCC	GMM-UBM	MFCC	LPCC	RASTA-PLP
White	0 dB	10.29	3.54	2.94	2.45	9.8
	5 dB	38.24	13.08	9.8	3.43	12.25
	10 dB	69.61	26.5	24.02	8.82	24.51
	15 dB	95.59	55.29	42.65	25	56.37
Pink	0 dB	9.31	10.67	16.67	7.35	10.29
	5 dB	45.1	21.92	28.92	15.69	24.51
	10 d	87.75	54.51	49.51	37.25	49.02
	15 d	95.59	88.09	86.27	72.55	91.18
Factory	0 dB	8.82	11.58	14.71	9.31	11.27
	5 dB	44.61	41.92	35.29	25	29.9
	10 d	87.75	60.04	66.18	52.94	63.24
	15 d	97.55	88.2	92.65	87.75	96.57
F16	0 dB	9.8	8.89	7.35	7.84	12.25
	5 dB	27.49	15.6	12.75	15.2	26.47
	10 d	69.12	45.63	52.94	36.76	50
	15 d	95.1	82.4	76.47	63.73	83.33