Buried Markov models for speech recognition
Author:
- 1 January 1999
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 2 (15206149) , 713-716 vol.2
Abstract
Good HMM-based speech recognition performance requires at most minimal inaccuracies to be introduced by HMM conditional independence assumptions. In this work, HMM conditional independence assumptions are relaxed in a principled way. For each hidden state value, additional dependencies are added between observation elements to increase both accuracy and discriminability. These additional dependencies are chosen according to natural statistical dependencies extant in training data that are not well modeled by an HMM. The result is called a buried Markov model (BMM) because the underlying Markov chain in an HMM is further hidden (buried) by specific cross-observation dependencies. Gaussian mixture HMMs are extended to represent BMM dependencies and new EM update equations are derived. On preliminary experiments with a large-vocabulary isolated-word speech database, BMMs are able to achieve an 11% improvement in WER with only a 9.5% increase in the number of parameters using a single state per mono-phone speech recognition system.Keywords
This publication has 6 references indexed in Scilit:
- Factorial Hidden Markov ModelsMachine Learning, 1997
- From HMM's to segment models: a unified view of stochastic modeling for speech recognitionIEEE Transactions on Speech and Audio Processing, 1996
- Graphical Models1996
- Connectionist Speech Recognition1994
- Could information theory provide an ecological theory of sensory processing?Network: Computation in Neural Systems, 1992
- A linear predictive HMM for vector-valued observations with applications to speech recognitionIEEE Transactions on Acoustics, Speech, and Signal Processing, 1990