Loading...
Please wait, while we are loading the content...
Similar Documents
Multi-stream asr trained with heterogeneous reverberant environments.
| Content Provider | CiteSeerX |
|---|---|
| Author | Shire, Michael |
| Abstract | A common problem with current automatic speech recognition (ASR) systems is that the performance degrades when it is presented with speech from a different acoustic environment than the one used during training. An important cause is that the feature distribution to which the ASR system is trained no longer matches that of a new environment. Reverberant environments can be especially harmful. In this work, we test a multi-stream system in which the constituent streams are each trained in separate acoustic environments. When training the acoustic modeling stages of the streams separately with clean data and heavily reverberated data, we find that that the combined system can improve the ASR performance with unseen reverberated test data. |
| File Format | |
| Access Restriction | Open |
| Subject Keyword | Asr Performance Acoustic Modeling Stage Clean Data Common Problem Feature Distribution Separate Acoustic Environment Different Acoustic Environment Multi-stream System Multi-stream Asr Trained New Environment Test Data Combined System Important Cause Heterogeneous Reverberant Environment Reverberant Environment Constituent Stream Asr System Current Automatic Speech Recognition |
| Content Type | Text |
