Abstract | ||
---|---|---|
We extend existing methods for automatic sentence boundary detection by leveraging multiple recognizer hypotheses in order to provide robustness to speech recognition errors. For each hypothesized word sequence, an HMM is used to estimate the posterior probability of a sentence boundary at each word boundary. The hypotheses are combined using confusion networks to determine the overall most likely events. Experiments show improved detection of sentences for conversational telephone speech, though results are mixed for broadcast news. |
Year | Venue | Keywords |
---|---|---|
2004 | HLT-NAACL (Short Papers) | multiple recognizer hypothesis,confusion network,automatic sentence boundary detection,sentence boundary,conversational telephone speech,improving automatic sentence boundary,speech recognition error,likely event,broadcast news,word sequence,word boundary,hypotheses,posterior probability,automatic,sequences,speech recognition,networks,boundaries |
Field | DocType | ISBN |
Broadcasting,Confusion,Computer science,Posterior probability,Robustness (computer science),Speech recognition,Boundary detection,Artificial intelligence,Natural language processing,Hidden Markov model,Sentence | Conference | 1-932432-24-8 |
Citations | PageRank | References |
8 | 0.80 | 7 |
Authors | ||
5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Dustin Hillard | 1 | 410 | 26.56 |
M. Ostendorf | 2 | 8 | 0.80 |
Andreas Stolcke | 3 | 6690 | 712.46 |
Y. Liu | 4 | 8 | 0.80 |
E. Shriberg | 5 | 10 | 2.18 |