Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2023
|
| Schlagworte: | |
| Online-Zugang: | https://arxiv.org/abs/2309.12802 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866917448770387968 |
|---|---|
| author | Ferreira, Alexandre R. Campelo, Cláudio E. C. |
| author_facet | Ferreira, Alexandre R. Campelo, Cláudio E. C. |
| contents | To train transcriptor models that produce robust results, a large and diverse labeled dataset is required. Finding such data with the necessary characteristics is a challenging task, especially for languages less popular than English. Moreover, producing such data requires significant effort and often money. Therefore, a strategy to mitigate this problem is the use of data augmentation techniques. In this work, we propose a framework that approaches data augmentation based on deepfake audio. To validate the produced framework, experiments were conducted using existing deepfake and transcription models. A voice cloner and a dataset produced by Indians (in English) were selected, ensuring the presence of a single accent in the dataset. Subsequently, the augmented data was used to train speech to text models in various scenarios. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2309_12802 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Deepfake audio as a data augmentation technique for training automatic speech to text transcription models Ferreira, Alexandre R. Campelo, Cláudio E. C. Sound Machine Learning Audio and Speech Processing I.2.6; I.2.0; E.0 To train transcriptor models that produce robust results, a large and diverse labeled dataset is required. Finding such data with the necessary characteristics is a challenging task, especially for languages less popular than English. Moreover, producing such data requires significant effort and often money. Therefore, a strategy to mitigate this problem is the use of data augmentation techniques. In this work, we propose a framework that approaches data augmentation based on deepfake audio. To validate the produced framework, experiments were conducted using existing deepfake and transcription models. A voice cloner and a dataset produced by Indians (in English) were selected, ensuring the presence of a single accent in the dataset. Subsequently, the augmented data was used to train speech to text models in various scenarios. |
| title | Deepfake audio as a data augmentation technique for training automatic speech to text transcription models |
| topic | Sound Machine Learning Audio and Speech Processing I.2.6; I.2.0; E.0 |
| url | https://arxiv.org/abs/2309.12802 |