ACL RD-TEC 1.0 Summarization of W02-0714
Paper Title:
IMPROVEMENTS IN NON-VERBAL CUE IDENTIFICATION USING MULTILINGUAL PHONE STRINGS
IMPROVEMENTS IN NON-VERBAL CUE IDENTIFICATION USING MULTILINGUAL PHONE STRINGS
Authors: Tanja Schultz and Qin Jin and Kornel Laskowski and Alicia Tribble and Alex Waibel
Primarily assigned technology terms:
- 3-way proficiency classification
- accent identification
- algorithm
- classification
- classification process
- classifier
- classifiers
- closed-set text-independent speaker identification
- coding
- computational linguistics
- cross-validation
- cue identification
- database
- decoding
- discriminant analysis
- error analysis
- evaluation procedure
- feature extraction
- hmm system
- identification
- identification process
- identification system
- language identification
- linear discriminant
- linear discriminant analysis
- lvcsr
- mean subtraction
- model training
- modeling
- n-gram modeling
- non-verbal cue identification
- normalization
- output coding
- perceptron
- phone recognition
- phone recognizer
- phonotactic model training
- proficiency classification
- reading
- recognition
- recognizer
- speaker identification
- speaker normalization
- speaker recognition
- speech-to-speech translation
- text-independent speaker identification
- transcription
- translation systems
- unsupervised training
- viterbi
- viterbi alignment
Other assigned terms:
- accent
- acoustic model
- acoustic models
- approach
- association for computational linguistics
- backoff
- case
- classification accuracy
- classification task
- concept
- confusion matrix
- correlation
- croatian
- data set
- data sets
- decision rule
- english language
- error rate
- evaluation set
- fact
- feature
- feature matrix
- french
- gaussian mixture
- gaussian mixture models
- generation
- hypothesis
- identification accuracy
- identification task
- index
- linguistics
- mandarin chinese
- methodology
- mixture models
- model parameter
- monolingual speech
- n-gram
- n-gram model
- n-gram models
- parameter space
- perplexity
- portability
- probability
- procedure
- process
- recognition component
- relation
- runtime
- russian
- segments
- speaking style
- spoken language
- style
- technique
- term
- test data
- testing data
- toolkit
- training
- training and testing data
- training data
- training set
- trigram
- user
- utterance
- vocabulary
- vocal tract
