Skip to main content
. Author manuscript; available in PMC: 2016 Dec 1.
Published in final edited form as: IEEE/ACM Trans Audio Speech Lang Process. 2016 Mar 1;24(5):967–977. doi: 10.1109/TASLP.2016.2536478

TABLE II.

STOI Comparison Between Speech Separation Methods With Single-SNR Speaker-Pair Dependent Training on SSC Corpus With Insufficient Clean Training Data

−12 dB −9 dB −6 dB −3 dB O dB 3 dB 6 dB
Noisy 46.4 52.2 58.8 65.8 72.7 79.2 84.8
DNN+Map 57.1 62.4 67.8 72.6 76.7 80.3 83.1
DNN+IRM 58.8 65.1 71.1 76.8 81.1 84.4 86.9
MCA+IRM 58.7 65.2 71.5 77.2 81.7 84.9 87.2
MCS+IRM 56.8 63.5 70.2 76.1 80.6 83.0 84.4
DNN+SA 66.0 71.6 77.0 81.4 85.2 88.2 90.5
MCA+SA 66.7 72.3 77.7 82.1 85.7 88.5 90.6
MCS+SA 66.7 72.3 77.7 82.0 85.4 88.1 90.1