Using Deep Neural Networks for Smoothing Pitch Profiles in Connected Speech
This paper presents a new pitch tracking smoother based on deep neural networks (DNN). It leverages Long Short-Term Memories, a particular kind of recurrent neural network, for correcting pitch detection errors produced by state-of-the-art Pitch Detection Algorithms. The proposed system has been ext...
Wedi'i Gadw mewn:
| Cyhoeddwyd yn: | URI:https://journals.openedition.org/ijcol, |
|---|---|
| Prif Awduron: | , |
| Fformat: | Article ou chapitre numérique |
| Iaith: | Anglais |
| Cyhoeddwyd: |
IJCoL
2020
|
| Mynediad Ar-lein: | Accès Université d'Orléans et IFPM Accès Université d'Orléans et IFPM |
| Crynodeb: | This paper presents a new pitch tracking smoother based on deep neural networks (DNN). It leverages Long Short-Term Memories, a particular kind of recurrent neural network, for correcting pitch detection errors produced by state-of-the-art Pitch Detection Algorithms. The proposed system has been extensively tested using two reference benchmarks for English and exhibited very good performances in correcting pitch detection algorithms outputs when compared with the gold standard obtained with laryngographs. |
|---|