A robust, segmental method for text independent speaker identification.
Herbert GishMichael SchmidtAngela MielkePublished in: ICASSP (1) (1994)
Keyphrases
- cost function
- computationally efficient
- computational cost
- high accuracy
- classification method
- segmentation method
- image noise
- genetic algorithm
- highly efficient
- synthetic data
- clustering method
- dynamic programming
- preprocessing
- objective function
- image processing
- classification accuracy
- significant improvement
- detection algorithm
- pairwise
- error rate
- matching algorithm
- bayesian networks
- high precision
- fully automatic
- similarity measure