Skip to main content

Advertisement

Table 3 Conditions for speaker recognition

From: Distant-talking speaker identification by generalized spectral subtraction-based dereverberation and its efficient computation

Sampling frequency 16 kHz
Frame length 25 ms
Frame shift 10 ms
Feature space 25 dimensions with CMN
  (12 MFCCs + Δ + Δ power)
Acoustic model GMMs with 128 diagonal
  covariance matrices