ACL RD-TEC 1.0 Summarization of H90-1037
Paper Title:
FAST SEARCH ALGORITHMS FOR CONNECTED PHONE RECOGNITION USING THE STOCHASTIC SEGMENT MODEL
FAST SEARCH ALGORITHMS FOR CONNECTED PHONE RECOGNITION USING THE STOCHASTIC SEGMENT MODEL
Authors: V.Digalakis and M. Ostendorf and J.R. Rohlicek
Primarily assigned technology terms:
- algorithm
- algorithm development
- classification
- classification algorithm
- classification method
- database
- dp search
- dynamic programming
- dynamic programming algorithm
- dynamic programming search
- fast search
- hidden markov
- hidden markov models
- hmms
- image segmentation
- maximum likelihood
- modelling
- phone classification
- phone recognition
- phonetic classification
- processing
- programming algorithm
- pruning
- recognition
- recognition algorithm
- recognition search
- recognizer
- score evaluation
- search
- search algorithm
- searching
- segment classification
- segmentation
- segmentation \
- segmentation algorithm
- shortest path
- speech processing
- splitting
- ssm recognition
- suboptimal search
Other assigned terms:
- approach
- bigram
- case
- classification accuracy
- classification scheme
- computational complexity
- concept
- conditional independence
- connectionist
- context information
- convergence
- correlation
- distribution
- duration
- evaluations
- experimental results
- fact
- feature
- feature vector
- feature vectors
- frame
- gaussian distribution
- grammar
- independence assumption
- likelihood
- likelihood ratio
- mapping
- markov models
- measure
- method
- model complexity
- oaussian distribution
- phoneme
- probabilities
- process
- recognition rate
- recursion
- search space
- search strategy
- segments
- sentence
- sentences
- statistical grammar
- stochastic segment model
- symbol
- term
- terms
- test set
- timit database
- training
- transformation
- tree
- unigram
