Abstract | ||
---|---|---|
In this paper we reports unsupervised training experiments we have conducted on large amounts of the English Fisher conversational telephone speech. A great amount of work has been reported on unsupervised training, but the major difference of this work is that we compared behaviors of unsupervised training with supervised training on exactly the same data. This comparison reveals surprising results. First, as the amount of training data increases, unsupervised training, even bootstrapped with a very limited amount (1 hour) of manual data, improves recognition performance faster than supervised training does, and it converges to supervised training. Second, bootstrapping unsupervised training with more manual data is not of significance if a large amount of un-transcribed data is available. |
Year | Venue | Keywords |
---|---|---|
2008 | INTERSPEECH 2008: 9TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION 2008, VOLS 1-5 | unsupervised training, supervised training, Fisher conversational telephone speech |
Field | DocType | Citations |
Computer science,Speech recognition,Unsupervised learning,Supervised training | Conference | 18 |
PageRank | References | Authors |
1.38 | 4 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jeff Z. Ma | 1 | 133 | 15.79 |
Richard M. Schwartz | 2 | 2839 | 717.76 |