A modified WaveNet for Filipino automatic speech recognition
In this study we explore the feasibility of using a modified WaveNet architecture for Automatic Speech Recognition (ASR) with the Filipino language. The ISIP corpus was used as a source of utterances. Two different models were trained. One used a subset of the training set that excluded prompts cons...
| Κύριος συγγραφέας: | |
|---|---|
| Μορφή: | Thesis |
| Γλώσσα: | English |
| Θέματα: |