ACL RD-TEC 1.0 Summarization of H92-1082
Paper Title:
AN EFFICIENT A* STACK DECODER ALGORITHM FOR CONTINUOUS SPEECH RECOGNITION WITH A STOCHASTIC LANGUAGE MODEL
AN EFFICIENT A* STACK DECODER ALGORITHM FOR CONTINUOUS SPEECH RECOGNITION WITH A STOCHASTIC LANGUAGE MODEL
Primarily assigned technology terms:
- acoustic matching
- algorithm
- caching
- computing
- continuous speech recognition
- continuous speech recognizer
- database
- decoder
- decoder algorithm
- fast match
- identification
- matching
- modeling
- network search
- phonetic modeling
- processing
- pruning
- recognition
- recognizer
- resource management
- scoring
- scoring function
- search
- search algorithm
- simulator
- speech recognition
- speech recognizer
- stack decoder
- top-n recognition
- tree search
- viterbi
- viterbi decoder
Other assigned terms:
- acoustic models
- bigram
- continuous speech
- cpu time
- derivation
- distribution
- evaluation function
- fact
- feature
- implementation
- language model
- language models
- leaf
- likelihood
- log-likelihood
- log-likelihood distribution
- method
- partial theory
- probabilities
- probability
- prototype implementation
- pruning threshold
- recognition task
- search problem
- sentence
- sentence fragment
- sentences
- speech input
- speech recognition problem
- stochastic language model
- technique
- theories
- theory
- tree
- trigram
- unigram
- unigram language model
- vocabulary
- vocabulary size
- word
- word sequence
- words
