ACL RD-TEC 1.0 Summarization of W02-0706
Paper Title:
ARCHITECTURES FOR SPEECH-TO-SPEECH TRANSLATION USING FINITE-STATE MODELS
ARCHITECTURES FOR SPEECH-TO-SPEECH TRANSLATION USING FINITE-STATE MODELS
Authors: Francisco Casacuberta and Enrique Vidal and Juan Miguel Vilar
Primarily assigned technology terms:
- algorithm
- approximation
- automatic speech recognition
- beam search
- beam-search
- computational linguistics
- computing
- decoding
- estimation process
- finite state
- finite state transducers
- finite-state technology
- finite-state transducer
- finite-state transducers
- finite-state translation
- hardware
- heuristic acceleration
- hidden markov
- hidden markov models
- hmms
- htk toolkit
- iterative procedure
- language modelling
- learning
- learning algorithms
- learning process
- learning technique
- maximum likelihood
- mgti technique
- modelling
- optimization
- pattern-recognition
- phonetic transcription
- recognition
- search
- search process
- searching
- segmentation
- sfst
- smoothing
- smoothing techniques
- spanish-to-english translation
- speech decoding
- speech recognition
- speech translation
- speech-to-speech translation
- stochastic finite-state transducer
- transcription
- transducer
- transducers
- translation process
- translation systems
- viterbi
- viterbi algorithm
- viterbi search
Other assigned terms:
- acoustic models
- approach
- association for computational linguistics
- beam
- break
- case
- convergence
- empty string
- english translation
- error rate
- estimation
- events
- fact
- finite-state grammar
- grammar
- heuristic
- hypothesis
- language model
- language models
- large training
- lexicon
- likelihood
- linguistics
- markov models
- n-gram
- optimization problem
- parallel text
- perplexity
- phoneme
- probability
- procedure
- process
- queries
- sentence
- sentences
- source language
- source sentence
- speech corpus
- statistical framework
- substring
- symbols
- target language
- target sentence
- target word
- technique
- technology
- term
- text
- toolkit
- training
- training corpus
- training data
- training set
- translation model
- translation models
- translation problem
- translations
- utterance
- word
- word error rate
- word sequence
- words
