Name
Title | ||
|---|---|---|
Minimum unit selection error training for HMM-based unit selection speech synthesis system |
Abstract | ||
|---|---|---|
This paper presents a minimum unit selection error (MUSE) training method for HMM-based unit selection speech synthesis system, which selects the optimal phone-sized unit sequence from the speech database by maximizing the combined likelihood of a group of trained HMMs. Under MUSE criterion, the weights and distribution parameters of these HMMs are estimated to minimize the number of different units between the selected phone sequences and the natural phone sequences for the training sentences. The optimization is realized by discriminative training using generalized probabilistic descent (GPD) algorithm. Results of our experiment show that this proposed method is able to improve the performance of the baseline system where model weights are set manually and distribution parameters are trained under maximum likelihood criterion. |
Year | DOI | Venue |
|---|---|---|
2008 | 10.1109/ICASSP.2008.4518518 | ICASSP |
Keywords | Field | DocType |
speech synthesis,generalized probabilistic descent algorithm,unit selection,maximum likelihood estimation,maximum likelihood criterion,hmm,minimum unit selection error,discriminative training,unit selection speech synthesis system,training sentences,speech database,baseline system,minimum unit selection error training,hidden markov models,phone sequences,probability,maximum likelihood | Speech synthesis,Pattern recognition,Computer science,Maximum likelihood,Speech recognition,Phone,Artificial intelligence,Probabilistic logic,Baseline system,Hidden Markov model,Discriminative model,Maximum likelihood criterion | Conference |
ISSN | ISBN | Citations |
1520-6149 E-ISBN : 978-1-4244-1484-0 | 978-1-4244-1484-0 | 3 |
PageRank | References | Authors |
0.45 | 3 | 2 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Zhen-Hua Ling | 1 | 850 | 83.08 |
| Ren-Hua Wang | 2 | 344 | 41.36 |