CONF aradilla:icassp:2007/IDIAP An Acoustic Model Based on Kullback-Leibler Divergence for Posterior Features Aradilla, Guillermo Vepa, Jithendra Bourlard, Hervé EXTERNAL https://publications.idiap.ch/attachments/papers/2007/aradilla-icassp-2007.pdf PUBLIC https://publications.idiap.ch/index.php/publications/showcite/aradilla:rr06-60 Related documents IEEE Int. Conf. on Acoustics, Speech, and Signal Processing (ICASSP) 2007 IDIAP-RR 06-60 This paper investigates the use of features based on posterior probabilities of subword units such as phonemes. These features are typically transformed when used as inputs for a hidden Markov model with mixture of Gaussians as emission distribution (HMM/GMM). In this work, we introduce a novel acoustic model that avoids the Gaussian assumption and directly uses posterior features without any transformation. This model is described by a finite state machine where each state is characterized by a target distribution and the cost function associated to each state is given by the Kullback-Leibler (KL) divergence between its target distribution and the posterior features. Furthermore, hybrid HMM/ANN system can be seen as a particular case of this KL-based model where state target distributions are predefined. A training method is also presented that minimizes the KL-divergence between the state target distributions and the posteriors features. REPORT aradilla:rr06-60/IDIAP An Acoustic Model Based on Kullback-Leibler Divergence for Posterior Features Aradilla, Guillermo Vepa, Jithendra Bourlard, Hervé EXTERNAL https://publications.idiap.ch/attachments/reports/2006/aradilla-idiap-rr-06-60.pdf PUBLIC Idiap-RR-60-2006 2006 IDIAP This paper investigates the use of features based on posterior probabilities of subword units such as phonemes. These features are typically transformed when used as inputs for a hidden Markov model with mixture of Gaussians as emission distribution (HMM/GMM). In this work, we introduce a novel acoustic model that avoids the Gaussian assumption and directly uses posterior features without any transformation. This model is described by a finite state machine where each state is characterized by a target distribution and the cost function associated to each state is given by the Kullback-Leibler (KL) divergence between its target distribution and the posterior features. Furthermore, hybrid HMM/ANN system can be seen as a particular case of this KL-based model where state target distributions are predefined. A training method is also presented that minimizes the KL-divergence between the state target distributions and the posteriors features.