New Features Using Robust MVDR Spectrum of Filtered Autocorrelation Sequence for Robust Speech Recognition

Table 1

Average recognition accuracies over different noise types and SNRs for test sets A, B, and C and different features.

Feature	Set A	Set B	Set C
MFCC	63.90	66.15	58.21
RAS-MFCC	68.10	69.06	63.70
PMCC	66.25	68.73	60.87
RPMCC	72.36	72.99	67.17
PMSR	70.28	71.11	65.79
R-PMSR	76.64	76.68	72.69