Abstract | ||
---|---|---|
Multi-stream and multi-band methods can improve the accu- racy of speech recognition systems without overly increasing the complexity. However, they cannot be applied blindly. In this pa- per, we review our experience applying multi-stream and multi- band methods to the Broadcast News corpus. We found that multi-stream systems using different acoustic front-ends provide a significant improvement over single stream systems. However, despite the fact that they have been successful on smaller tasks, we have not yet been able to show any improvement using multi- band methods. We report various insights gained from the expe- rience in applying these methods in a large-vocabulary task. |
Year | Venue | Keywords |
---|---|---|
1999 | EUROSPEECH | front end,speech recognition,artificial intelligence,electrical engineering |
Field | DocType | Citations |
Broadcasting,Prime time,Speech analytics,Computer science,Speech recognition,Speech technology | Conference | 27 |
PageRank | References | Authors |
7.13 | 4 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Adam Janin | 1 | 250 | 34.11 |
Daniel P. W. Ellis | 2 | 4198 | 356.08 |
Nelson Morgan | 3 | 3048 | 533.52 |