Figure 3From: Integrated exemplar-based template matching and statistical modeling for continuous speech recognition Using three PDT clustering structures to calculate the template matching score. Back to article page