Abstract | ||
---|---|---|
In this paper, we describe the BBN 2007 Mandarin Speech-to-Text system developed for the GALE Evaluation 2007. In comparison to the BBN 2006 Mandarin system, we achieved 25% relative reduction in character error rate on the most important test sets. The utilization of all available training data provided the largest contribution to the improvement. The use of a better pitch tracking algorithm also contributed significantly, while system combination made some noticeable improvement too. |
Year | DOI | Venue |
---|---|---|
2008 | 10.1109/ICASSP.2008.4517915 | ICASSP |
Keywords | Field | DocType |
natural languages,speech recognition,BBN 2007 Mandarin speech-to-text system,GALE Evaluation 2007,pitch tracking algorithm,speech recognition,Mandarin,Speech recognition,pitch,system combination | Training set,System combination,Computer science,Word error rate,Speech recognition,Natural language,Artificial intelligence,Natural language processing,Pitch tracking,Mandarin Chinese | Conference |
ISSN | Citations | PageRank |
1520-6149 | 9 | 0.72 |
References | Authors | |
5 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Truong Ng | 1 | 9 | 0.72 |
Bing Zhang | 2 | 13 | 1.56 |
Kham Nguyen | 3 | 47 | 3.64 |
Long Nguyen | 4 | 326 | 84.60 |