A Minimum Cross-Entropy Approach to Hidden Markov Model Adaptation - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Journal Articles IEEE Signal Processing Letters Year : 1999

A Minimum Cross-Entropy Approach to Hidden Markov Model Adaptation

Yifan Gong
  • Function : Author
Jean-Paul Haton
  • Function : Author
  • PersonId : 830987

Abstract

An adaptation algorithm using the theoretically optimal maximum a posteriori (MAP) formulation, and at the same time accounting for parameter correlation between different classes is desirables, especially when using sparse adaptation data. However, a direct implementation of such an approach may be prohibitive in many practical situations. In this letter, we present an algorithm that approximates the above mentioned correlated MAP algorithm by iteratively maximizing the set of posterior marginals. With some simplifying assumptions, expression for these marginals are then derived, using the principle of minimum cross-entropy. The resulting algorithm is simple, and includes conventional MAP estimation as a special case. The utility of the proposed method is tested in adaptation experiments for an alphabet recognition task.
No file

Dates and versions

inria-00098969 , version 1 (26-09-2006)

Identifiers

  • HAL Id : inria-00098969 , version 1

Cite

Mohamed Afify, Yifan Gong, Jean-Paul Haton. A Minimum Cross-Entropy Approach to Hidden Markov Model Adaptation. IEEE Signal Processing Letters, 1999, 6 (6), pp.132-134. ⟨inria-00098969⟩
95 View
0 Download

Share

Gmail Facebook Twitter LinkedIn More