ACL RD-TEC 1.0 Summarization of H92-1035
Paper Title:
IMPROVING STATE-OF-THE-ART CONTINUOUS SPEECH RECOGNITION SYSTEMS USING THE N-BEST PARADIGM WITH NEURAL NETWORKS
IMPROVING STATE-OF-THE-ART CONTINUOUS SPEECH RECOGNITION SYSTEMS USING THE N-BEST PARADIGM WITH NEURAL NETWORKS
Authors: S. Austin and G. Zavaliagkos and J. Makhoul and R. Schwartz
Primarily assigned technology terms:
- 1-best training
- acoustic modeling
- algorithm
- continuous speech recognition
- discriminative training
- hidden markov
- hidden markov models
- hmm csr
- hmm system
- hmms
- language processing
- large vocabulary speech recognition
- linear interpolation
- matching
- modeling
- n-best paradigm
- n-best paradigm \
- n-best rescoring
- n-best training
- natural language processing
- neural net
- neural network
- neural networks
- phoneme insertion
- phonetic modeling
- processing
- recognition
- recognition system
- recognition systems
- recognizer
- rescoring
- resource management
- sampling
- scoring
- search
- search technique
- searching
- segmental neural nets
- segmentation
- speech recognition
- speech recognition \
- speech recognition system
- speech recognition systems
- time warping
- time-waiping function
- training algorithm
- training method
- transcription
Other assigned terms:
- byblos system
- case
- concept
- continuous speech
- correlation
- development set
- duration
- error rate
- feature
- feature vectors
- frame
- grammar
- grammars
- histogram
- hypotheses
- hypothesis
- interpolation
- knowledge
- large vocabulary speech
- likelihood
- linear combination
- markov models
- measure
- method
- model context
- n-best list
- natural language
- perplexity
- phoneme
- phonemes
- probabilities
- probability
- process
- recognition component
- recognition task
- segments
- sentence
- sentences
- sigmoid function
- speech corpus
- speech recognition performance
- technique
- technologies
- term
- test set
- training
- training criterion
- training data
- training set
- transcriptions
- utterance
- vocabulary
- word
- word error rates
- word sequence
- word-pair grammar
- words
