Multimode Speech Coding - Digital Speech: Coding for Low Bit Rate Communication Systems - page 333

9.11.1 PerformanceUnderAcousticNoise
The classification algorithm was tested using 64 seconds of male and female
speech corrupted with either babble or vehicular noise. The SNR of the
corrupted speech is 10 dB.
Figure 9.31 depicts the classification of the female speech. The initial clas-
sification declares only the strongly-unvoiced segments as unvoiced and all

AHHHHHHAA

0 500 1000 1500

samples
(a) Stationary voiced speech

HAAANN A A A

0 500 1000 1500

samples

(b) Unvoiced speech

Figure 9.31 Classification of female speech corrupted by babble noise (10 dB SNR):
A (ACELP), H (harmonic), and N (noise excitation)