Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2404.00863 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909760336429056 |
|---|---|
| author | Tao, Ruijie Shi, Zhan Jiang, Yidi Liu, Tianchi Li, Haizhou |
| author_facet | Tao, Ruijie Shi, Zhan Jiang, Yidi Liu, Tianchi Li, Haizhou |
| contents | Modern speaker recognition system relies on abundant and balanced datasets for classification training. However, diverse defective datasets, such as partially-labelled, small-scale, and imbalanced datasets, are common in real-world applications. Previous works usually studied specific solutions for each scenario from the algorithm perspective. However, the root cause of these problems lies in dataset imperfections. To address these challenges with a unified solution, we propose the Voice Conversion Augmentation (VCA) strategy to obtain pseudo speech from the training set. Furthermore, to guarantee generation quality, we designed the VCA-NN~(nearest neighbours) strategy to select source speech from utterances that are close to the target speech in the representation space. Our experimental results on three created datasets demonstrated that VCA-NN effectively mitigates these dataset problems, which provides a new direction for handling the speaker recognition problems from the data aspect. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_00863 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Voice Conversion Augmentation for Speaker Recognition on Defective Datasets Tao, Ruijie Shi, Zhan Jiang, Yidi Liu, Tianchi Li, Haizhou Audio and Speech Processing Modern speaker recognition system relies on abundant and balanced datasets for classification training. However, diverse defective datasets, such as partially-labelled, small-scale, and imbalanced datasets, are common in real-world applications. Previous works usually studied specific solutions for each scenario from the algorithm perspective. However, the root cause of these problems lies in dataset imperfections. To address these challenges with a unified solution, we propose the Voice Conversion Augmentation (VCA) strategy to obtain pseudo speech from the training set. Furthermore, to guarantee generation quality, we designed the VCA-NN~(nearest neighbours) strategy to select source speech from utterances that are close to the target speech in the representation space. Our experimental results on three created datasets demonstrated that VCA-NN effectively mitigates these dataset problems, which provides a new direction for handling the speaker recognition problems from the data aspect. |
| title | Voice Conversion Augmentation for Speaker Recognition on Defective Datasets |
| topic | Audio and Speech Processing |
| url | https://arxiv.org/abs/2404.00863 |