ACL RD-TEC 1.0 Summarization of H94-1087
Paper Title:
LANGUAGE IDENTIFICATION VIA LARGE VOCABULARY SPEAKER INDEPENDENT CONTINUOUS SPEECH RECOGNITION
LANGUAGE IDENTIFICATION VIA LARGE VOCABULARY SPEAKER INDEPENDENT CONTINUOUS SPEECH RECOGNITION
Authors: Steve Lowe and Anne Demedts and Larry Gillick and Mark Mandel and Barbara Peskin
Primarily assigned technology terms:
- acoustic normalization
- approximation
- blind deconvolution
- classification
- computing
- continuous speech recognition
- data collection
- deconvolution
- discriminant analysis
- hidden markov
- hidden markov model
- hidden markov models
- hmms
- identification
- identification system
- independent continuous speech recognition
- language classification
- language identification
- large vocabulary continuous speech recognition
- linear discriminant
- linear discriminant analysis
- lvcsr
- markov model
- message classification
- mixture modelling
- modelling
- normalization
- phone recognition
- recognition
- recognition system
- recognition systems
- recognizer
- rescoring
- score normalization
- scoring
- signal-processing
- speaker identification
- speaker-independent recognition
- speech recognition
- speech recognition system
- speech recognition systems
- speech recognizer
- stochastic process
- topic identification
- topic-independent recognizer
- transcription
Other assigned terms:
- acoustic models
- acoustic score
- acoustic signal
- approach
- bias
- case
- classification problem
- classification tasks
- conditional probability
- contextual information
- continuous speech
- conversational telephone speech
- data consortium
- distribution
- duration
- english telephone speech
- events
- fact
- feature
- feature vector
- feature vectors
- frame
- full speech recognition
- grammar
- hypotheses
- hypothesis
- implementation
- joint probability
- language discrimination
- language model
- language models
- likelihood
- likelihood ratio
- linguistic
- linguistic data
- linguistic data consortium
- markov models
- measures
- message
- monologue
- natural speech
- parameter space
- pauses
- phoneme
- phonemes
- prior probability
- probabilities
- probability
- probability distributions
- process
- pronunciation
- recognition quality
- recognition score
- recognition task
- segments
- sentences
- signal
- spanish language
- speaking rate
- speech data
- stochastic model
- switchboard corpus
- term
- terms
- test corpus
- test data
- test material
- text
- theoretical framework
- theory
- topics
- training
- training data
- training material
- transcript
- transcriptions
- transformation
- utterance
- vocabulary
- word
- word pronunciation
- word sequence
- words
