ACL RD-TEC 1.0 Summarization of H90-1004
Paper Title:
A TREE.TRELLIS BASED FAST SEARCH FOR FINDING THE N BEST SENTENCE HYPOTHESES IN CONTINUOUS SPEECH RECOGNITION
A TREE.TRELLIS BASED FAST SEARCH FOR FINDING THE N BEST SENTENCE HYPOTHESES IN CONTINUOUS SPEECH RECOGNITION
Authors: Frank K. Soong and Eng-Fong Huang
Primarily assigned technology terms:
- algorithm
- backtracking
- beam search
- beam-search
- best-first search
- breadth-first search
- computing
- continuous speech recognition
- continuous speech recognizer
- data base
- database
- decoding
- digit recognition
- dynamic programming
- error detection
- fast search
- fast tree-trellis
- finite state
- finite state automata
- graph search
- hmms
- language processing
- language understanding
- n-best search
- natural language processing
- network search
- parser
- path growing
- processing
- ranking
- recognition
- recognition system
- recognizer
- resource management
- search
- search algorithm
- segmentation
- semantic analysis
- sentence parser
- sentence segmentation
- speech recognition
- speech recognition system
- speech recognizer
- spoken language system
- spoken language understanding
- stack-based decoding
- telecommunication
- tree search
- tree-trellis search
- tree-treuis search
- viterbi
- viterbi algorithm
- viterbi decoding
- viterbi search
Other assigned terms:
- accent
- acoustic information
- approach
- automata
- beam
- computational overhead
- content words
- continuous speech
- cpu time
- data structures
- density function
- finite state grammar
- finite state network
- frame
- gaussian mixture
- grammar
- heuristic
- heuristics
- hypotheses
- hypothesis
- implementation
- index
- input string
- knowledge
- likelihood
- measure
- natural language
- noise
- perplexity
- probability
- probability density
- probability density function
- procedure
- recognition application
- resource management task
- root node
- search problem
- search procedure
- search space
- search tree
- semantic
- sentence
- sentences
- silence
- spoken language
- symbols
- terms
- test data
- test set
- tokens
- tone
- training
- training data
- tree
- understanding
- vocabulary
- word
- word boundaries
- word-pair grammar
- words
