ACL RD-TEC 1.0 Summarization of H93-1013
Paper Title:
SESSION 3: CONTINUOUS SPEECH RECOGNITION
SESSION 3: CONTINUOUS SPEECH RECOGNITION
Primarily assigned technology terms:
- acoustic modeling
- adaptive language modeling
- caching
- cepstral normalization
- continuous speech recognition
- fast search
- forward search
- identification
- keyword spotting
- language identification
- language modeling
- large vocabulary continuous speech recognition
- large vocabulary speech recognition
- large-vocabulary continuous speech recognition
- maximum entropy
- modeling
- n-best search
- normalization
- progressive-search
- real-time recognition
- recognition
- recognition system
- recognition-time adaptation
- recognizer
- robust speech recognition
- search
- searching
- segment recognition
- speaker identification
- speech recognition
- speech recognition system
- sphinx-ii
- topic identification
Other assigned terms:
- acoustic models
- additive noise
- case
- continuous speech
- entropy
- feature
- feature set
- information sources
- keyword
- language model
- language models
- large vocabulary speech
- maximum entropy principle
- method
- noise
- perplexity
- probabilities
- probability
- search space
- search strategy
- sentence
- signal
- speaker identity
- technique
- vocabulary
- word
- word sequence
- words
