Feature Extraction Approach for Speaker Verification to Support Healthcare System Using Blockchain Security for Data Privacy

Table 4

Error rate in less acoustic scenario (noise).

Dialects	Train	Test	Feature extractions
Hindi	(sec)	(sec)	LPC (%ER)	MFCC (%ER)	LPCC (%ER)
lnk	10	10	10.01	8.01	9.11
/kU; okn	10	5	11.88	7.77	10.02
gkWa	10	20	9.90	10.03	11.73
okyk	5	10	12.74	10.75	11.88
ebZ	5	5	9.88	8.75	9.78
vdcj	5	20	7.71	9.77	8.33
iz t	20	10	9.46	6.56	7.19
ia[kk	20	5	10.89	9.77	7.36
eka; k	20	20	9.22	5.90	9.79
vua	20	10	8.66	6.57	12.34