Table 7 Results using two CD codebooks and on CI codebook, for context match scenario for CD-1 under babble noise
From: Robust Bayesian estimation for context-based speech enhancement
|  | ΔPESQ | ΔSSNR (in dB) | ||||
|---|---|---|---|---|---|---|
Input SNR | 0 dB | 5 dB | 10 dB | 0 dB | 5 dB | 10 dB |
CBSE with CD-1 | 0.18±0.14 | 0.18±0.16 | 0.12±0.14 | 5.87±1.16 | 4.88±1.14 | 2.81±1.27 |
CBSE with CD-2 | 0.08±0.18 | 0.05±0.16 | −0.03±0.12 | 5.69±1.30 | 4.52±1.17 | 2.18±1.31 |
CBSE with CI | 0.04±0.17 | 0.02±0.15 | −0.11±0.17 | 5.41±1.20 | 4.39±1.16 | 1.98±1.28 |
Proposed | 0.17±0.13 | 0.16±0.14 | 0.07±0.16 | 5.81±1.14 | 4.87±1.10 | 2.58±1.33 |
Reference | −0.03±0.07 | −0.03±0.06 | −0.07±0.07 | 1.76±0.50 | 1.93±0.43 | 1.66±0.57 |