Figure 5
From: Decision tree-based acoustic models for speech recognition
Distribution of feature usage (vowel classes) in the DTAM triphone system without gender information.
Figure 5 | EURASIP Journal on Audio, Speech, and Music Processing