Segment-Based Acoustic Models for Continuous Speech Recognition
Abstract
In work, we are interested in the problem of large vocabulary, speaker-independent continuous speech recognition, and primarily in the acoustic modeling component of this problem. In developing acoustic models for speech recognition, we have conflicting goals. On one hand, the models should be robust to inter- and intra-speaker variability, to the use of a different vocabulary in recognition than in training, and to the effects of moderately noisy environments. In order to accomplish this, we need to model gross features and global trends. On the other hands, the models must be sensitive and detailed enough to detect fine acoustic differences between similar words in a large vocabulary task. To answer these opposing demands requires improvements in acoustic modeling at several levels: the frame level (e.g. signal processing), the phoneme level (e.g. modeling feature dynamics), and the utterance level (e. g. defining a structural context for representing the intra-utterance dependence across phonemes). This project address the problem of acoustic modelling specifically focusing on modeling at the segment level and above
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 11, 1994
- Accession Number
- ADA276109
Entities
People
- J. R. Rohlicek
- Mari Ostendorf
Organizations
- Boston University