ACL RD-TEC 1.0 Summarization of H94-1086
Paper Title:
ON-LINE CURSIVE HANDWRITING RECOGNITION USING HIDDEN MARKOV MODELS AND STATISTICAL GRAMMARS
ON-LINE CURSIVE HANDWRITING RECOGNITION USING HIDDEN MARKOV MODELS AND STATISTICAL GRAMMARS
Authors: John Makhoul and Thad Starner and Richard Schwartz and George Chou
Primarily assigned technology terms:
- atis
- character recognition
- continuous speech recognition
- data collection
- decoder
- digit recognition
- handwriting recognition
- hardware
- hidden markov
- hidden markov models
- hmms
- language modeling
- modeling
- on-line handwriting recognition
- optimization
- preprocessing
- recognition
- recognition system
- recognition system \
- recognizer
- sampling
- segmentation
- speech recognition
- speech recognition system
- speech system
- transcription
- tuning
Other assigned terms:
- automatic optimization
- bigram
- byblos system
- case
- character error rate
- characters
- clusters
- continuous speech
- corpora
- error rate
- feature
- feature set
- feature vector
- feature vectors
- grammar
- grammars
- grid
- handwriting
- hmm model
- language models
- lexicon
- markov models
- method
- noise
- pairs of words
- perplexity
- phonemes
- process
- punctuation
- real-time performance
- recognition accuracy
- scalability
- sentence
- sentences
- signal
- statistical grammar
- symbol
- symbols
- test data
- test set
- text
- time series
- training
- training text
- transcriptions
- vocabulary
- word
- word boundaries
- word error rate
- word error rates
- words
