Skip to main content

Table 5 Description of each speaker recognition method

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

Method Speaker Processing at
number models test stage
1 (Baseline) Clean speech models CMN with
2 (Method in [22]) Clean speech models GSS-based
3 Reverberant CMN with
  speech models beamforming
4 (MSLP-based method) Dereverberant MSLP-GSS-based
  speech models  
  based on MSLP-GSS dereverberation
5 (Proposed method) Dereverberant MCLMS-GSS-based
  speech models  
  based on MCLMS-GSS dereverberation