Paper available on hidden Markov model inversion

[Seokyong Moon]

FTP-host:     pierce.ee.washington.edu
FTP-filename: /pub/papers/hmm-inversion.ps.Z

This paper is 30 pages long.

	Robust Speech Recognition using Gradient-Based Inversion 
	   and Baum-Welch Inversion of Hidden Markov Models

		Seokyong Moon, Jenq-Neng Hwang

	Information Processing Laboratory
	Department of Electrical Engineering, FT-10
	University of Washington, Seattle, WA 98195
	E-mail: moon at pierce.ee.washington.edu, hwang at ee.washington.edu

The gradient based hidden Markov model (HMM) inversion algorithm is studied
and applied to robust speech recognition tasks under general types of
mismatched conditions.  It stems from the gradient-based inversion
algorithm of an artificial neural network (ANN) by viewing an HMM as a
special type of ANNs. The HMM inversion has a conceptual duality to the HMM
training just as ANN inversion does to ANN training.  The forward training
of an HMM, based on either the Baum-Welch reestimation or gradient method,
finds the model parameters to optimize some criteria, e.g., maximum
likelihood (ML), maximum mutual information (MMI) and mean squared error
(MSE), with given speech inputs. On the other hand, the inversion of an HMM
finds speech inputs that optimize some criterion with given model
parameters.  The performance of the proposed gradient based HMM inversion
for noisy speech recognition under additive noise corruption and microphone
mismatch conditions is compared with the robust Baum-Welch HMM inversion
technique along with other noisy speech recognition technique, i.e., the
robust minimax (MINIMAX) classification technique.