Speaker independent recognition of isolated words using clustering techniques
- 24 March 2005
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 4, 574-577
- https://doi.org/10.1109/icassp.1979.1170821
Abstract
A speaker independent, isolated word recognition system is proposed which is based on the use of multiple templates for each word in the vocabulary. The word templates are obtained from a statistical clustering analysis of a large data base consisting of 100 replications of each word (i.e. once by each of 100 talkers). The recognition system, which uses telephone recordings, is based on an LPC analysis of the unknown word, dynamic time warping of each reference template to the unknown word (using the Itakura LPC distance measure), and the application of a K-nearest neighbor (KNN) decision rule to lower the probability of error. Results are presented on two test sets of data which show error rates that are comparable to, or better than, those obtained with speaker trained, isolated word recognition systems.Keywords
This publication has 4 references indexed in Scilit:
- Considerations in dynamic time warping algorithms for discrete word recognitionIEEE Transactions on Acoustics, Speech, and Signal Processing, 1978
- Dynamic programming algorithm optimization for spoken word recognitionIEEE Transactions on Acoustics, Speech, and Signal Processing, 1978
- On creating reference templates for speaker independent recognition of isolated wordsIEEE Transactions on Acoustics, Speech, and Signal Processing, 1978
- Minimum prediction residual principle applied to speech recognitionIEEE Transactions on Acoustics, Speech, and Signal Processing, 1975