Show simple item record

dc.contributor.authorRenals, Steveen
dc.contributor.authorMorgan, Nelsonen
dc.contributor.authorCohen, Michaelen
dc.contributor.authorFranco, Horacioen
dc.coverage.spatial4en
dc.date.accessioned2006-05-18T17:26:16Z
dc.date.available2006-05-18T17:26:16Z
dc.date.issued1992-03
dc.identifier.citationAcoustics, Speech, and Signal Processing, 1992. ICASSP-92., 1992 IEEE International Conference on, Volume 1, 23-26 March 1992 Page(s):601 - 604.
dc.identifier.issn1520-6149
dc.identifier.uriDigital Object Identifier 10.1109/ICASSP.1992.225837
dc.identifier.urihttp://ieeexplore.ieee.org/
dc.identifier.urihttp://hdl.handle.net/1842/1131
dc.description.abstractThe authors have previously demonstrated that feedforward networks can be used to estimate local output probabilities in hidden Markov model (HMM) speech recognition systems (Renals et al., 1991). These connectionist techniques are integrated into the DECIPHER system, with experiments being performed using the speaker-independent DARPA RM database. The results indicate that: connectionist probability estimation can improve performance of a context-independent maximum-likelihood-trained HMM system; performance of the connectionist system is close to what can be achieved using (context-dependent) HMM systems of much higher complexity; and mixing connectionist and maximum-likelihood estimates can improve the performance of the state-of-the-art context-independent HMM system.en
dc.format.extent383305 bytesen
dc.format.mimetypeapplication/pdfen
dc.language.isoen
dc.publisherIEEEen
dc.titleConnectionist probability estimation in the DECIPHER speech recognition systemen
dc.typeConference Paperen


Files in this item

This item appears in the following Collection(s)

Show simple item record