Segment-Based Acoustic Models for Continuous Speech Recognition
Abstract
This research aims to develop new and more accurate stochastic models for speaker-independent continuous speech recognition, by extending previous work in segment-based modeling and by introducing a new hierarchical approach to representing intra-utterance statistical dependencies. These techniques, which are more costly than traditional approaches because of the large search space associated with higher order models, are made feasible through rescoring a set of HMM-generated N-best sentence hypotheses. We expect these different modeling techniques to result in improved recognition performance over that achieved by current systems, which handle only frame-based observations and assume that these observations are independent given an underlying state sequence. In the fourth quarter of the project, we have: (1) ported our recognition system to the Wall Street Journal task, a standard task in the ARPA community; (2) developed an initial dependency-tree model of intra-utterance observation correlation; and (3) implemented baseline language model estimation software. Our initial results on the Wall Street Journal task are quite good, representing improved performance over most HMM systems reporting on the November 1992 5k vocabulary test set.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 08, 1993
- Accession Number
- ADA267138
Entities
People
- J. R. Rohlicek
- Mari Ostendorf
Organizations
- Boston University