Skip to main content

Advertisement

Table 1 Speaker identification baseline for each type of simulated (noisy) reverberant data

From: Single-channel dereverberation by feature mapping using cascade neural networks for robust distant speaker identification and speech recognition

RIR Speaker identification rate (%)
Type T 60  (ms) Clean Reverb. Noisy reverberant
SNR =  SNR = 20 dB SNR = 10 dB
Office 400 98.0 74.3 55.3 20.6
Livingroom 70.9 - -