Iterated Class-Specific Subspaces for Speaker-Dependent Phoneme Classification
Abstract
The features based on the MEL cepstrum have long dominated probabilistic methods in automatic speech recognition (ASR). This feature set has evolved to maximize general ASR performance within a Bayesian classifier framework using a common feature space. Now, however, with the advent of the PDF projection theorem (PPT) and the class-specific method (CSM), it is possible to design features separately for each phoneme and compare log-likelihood values fairly across various feature sets. In this paper, class-dependent features are found by optimizing a set of frequency-band functions for projection of the spectral vectors, analogous to the MEL frequency band functions, individually for each class. Using this method, we show significant improvement over standard MEL cepstrum methods in speaker and phoneme specific recognition.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2008
- Accession Number
- ADA494622
Entities
People
- Paul Baggenstoss
Organizations
- Naval Undersea Warfare Center