Integrated Stochastic Pronunciation Modeling. Dong Wang Supervisors: Simon King, Joe Frankel, James Scobbie. Contents. Problems we are addressing Previous research Integrated stochastic pronunciation modeling Current experimental results Work plan. Problems we are addressing.
Supervisors: Simon King, Joe Frankel, James Scobbie
1. Linguistics-based lexica
2. determinate mapping
Integrated Stochastic Pronunciation Modeling (ISPM)
Direct grapheme ISPM: SSM is a 1:1 mapping
Hidden grapheme ISPM: SSM is a n:m mapping
Performance gain expectation from ISPM
Data corpora for the LVSR task
WSJCAM0 for read speech and RT04S for spontaneous speech on the meeting domain
Experiment settings for the LVSR task
Experimental results of the LVSR task
Contribution of context dependent modeling
Contribution of phonology oriented questions to the grapheme system
sub-word lattice based architecture for STD
STD performance on the RT04S task
We have demonstrated that in Spanish, which holds simple grapheme-phoneme relationship and achieves close ASR performance with phoneme and grapheme based systems, the grapheme-based STD system outperforms the phoneme-based one.
parallel phone/grapheme recognizer architecture for LID
1. Finish the STD experiments with high-order LMs (by Jan.2008).
2. Finish the LID oriented tuning (by Nov.2007).
3. Apply powerful LMs to the LID task (by Jan.2008).
4. Finish the SSM design (by Jan.2008).
5. Apply the SSM on LVSR RTS04 and STD (by Feb.2008).
1. Finish the direct-grapheme architecture (GPM) design (by Jul.2008).
2. Test the direct-grapheme architecture on the LVSR RTS04 task (by Oct.2008).
3. Finish the hidden-grapheme architecture (GPM+SSM) (by Jan.2009).
4. Test the hidden-grapheme architecture on the LVSR RTS04 task (by Feb.2009).
1. Finish the test on the STD task (by May 2009).
2. Finish the test on the LID task (by May 2009).