ACL RD-TEC 1.0 Summarization of H92-1066
Paper Title:
RAPID MATCH TRAINING FOR LARGE VOCABULARIES
RAPID MATCH TRAINING FOR LARGE VOCABULARIES
Authors: Larry Gillick and Barbara Peskin and Robert Roth
Primarily assigned technology terms:
- algorithm
- alignment process
- clustering
- clustering algorithm
- continuous speech recognition
- continuous speech recognizer
- hidden markov
- hidden markov models
- hmms
- linear alignment
- reading
- recognition
- recognition process
- recognizer
- sampling
- scoring
- segmentation
- speech recognition
- speech recognizer
- spelling
- training method
- training procedure
- training process
- transcription
- tuning
- word recognizer
Other assigned terms:
- acoustic information
- case
- cluster
- clusters
- continuous speech
- correlation
- distribution
- duration
- error rate
- fact
- frame
- function words
- hypotheses
- implementation
- language model
- language model score
- likelihood
- markov models
- measure
- method
- model parameters
- parameter settings
- parameter values
- phoneme
- probability
- probability density
- probability distributions
- procedure
- process
- punctuation
- recognition task
- representations
- run-time
- sentence
- sentences
- silence
- speech data
- test corpus
- test set
- text
- tokens
- training
- utterance
- vocabulary
- vocabulary size
- wall street journal corpus
- word
- word candidate
- word error rates
- word model
- words
