13th Annual Conference of the International Speech Communication Association

Portland, OR, USA
September 9-13, 2012

Robust Triphone Mapping for Acoustic Modeling

Miloš Cerňak (1,2), David Imseng (1,3), Hervé Bourlard (1,3)

(1) Idiap Research Institute, Martigny, Switzerland
(2) Institute of Informatics, Slovak Academy of Sciences, Bratislava, Slovakia
(3) Ecole Polytechnique Fédérale, Lausanne (EPFL), Switzerland

In this paper we revisit the recently proposed triphone mapping as an alternative to decision tree state clustering. We generalize triphone mapping to Kullback-Leibler based hidden Markov models for acoustic modeling and propose a modified training procedure for the Gaussian mixture model based acoustic modeling. We compare the triphone mapping to decision tree state clustering on the Wall Street journal task as well as in the context of an under-resourced language by using Greek data from the SpeechDat(II) corpus. Experiments reveal that triphone mapping has the best overall performance and is robust against varying the acoustic modeling technique as well as variable amounts of training data.

Index Terms: Speech recognition, acoustic modeling, triphone mapping, Kullback-Leibler divergence

Full Paper

Bibliographic reference.  Cerňak, Miloš / Imseng, David / Bourlard, Hervé (2012): "Robust triphone mapping for acoustic modeling", In INTERSPEECH-2012, 1910-1913.