Skip to main content

Advertisement

Table 6 Distant-talking speaker recognition rates of artificial data (%)

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

  Number of impulse response  
Method number condition for test Average
4 5 6 7 8
1 66.7 53.3 43.2 43.7 38.3 49.0
2 53.1 32.9 25.6 25.3 29.1 33.2
3 91.6 88.4 86.5 87.6 88.0 88.4
4 93.7 90.2 89.8 89.9 89.2 90.6
5 94.0 90.6 91.0 90.5 92.3 91.7