From: Front-end technologies for robust ASR in reverberant environments—spectral enhancement-based dereverberation and auditory modulation filterbank features
f s
STFT block
ξ min (5)
τ s (10)
L e (10)
(μ,γ,p 0,p ∞ ) (16)-(18)
G min (19)
K d (20)
16 kHz
32 ms
−30 dB
16 ms
3
(0.5,0.5,0.5,1.0)
−10 dB
80
