Fig. 7From: A Bayesian view on acoustic model-based techniques for robust speech recognitionBayesian network representation of reverberant VTS (Subsection 6.6.1) a before and b after approximation via an extended observation vector. The figure is based on [15]Back to article page