Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2506.17690 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918067395624960 |
|---|---|
| author | Herreilers, Julian Jacobs, Christiaan Niesler, Thomas |
| author_facet | Herreilers, Julian Jacobs, Christiaan Niesler, Thomas |
| contents | We introduce a new approach, the ContrastiveTransformer, that produces acoustic word embeddings (AWEs) for the purpose of very low-resource keyword spotting. The ContrastiveTransformer, an encoder-only model, directly optimises the embedding space using normalised temperature-scaled cross entropy (NT-Xent) loss. We use this model to perform keyword spotting for radio broadcasts in Luganda and Bambara, the latter a severely under-resourced language. We compare our model to various existing AWE approaches, including those constructed from large pre-trained self-supervised models, a recurrent encoder which previously used the NT-Xent loss, and a DTW baseline. We demonstrate that the proposed contrastive transformer approach offers performance improvements over all considered existing approaches to very low-resource keyword spotting in both languages. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_17690 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Low-resource keyword spotting using contrastively trained transformer acoustic word embeddings Herreilers, Julian Jacobs, Christiaan Niesler, Thomas Audio and Speech Processing We introduce a new approach, the ContrastiveTransformer, that produces acoustic word embeddings (AWEs) for the purpose of very low-resource keyword spotting. The ContrastiveTransformer, an encoder-only model, directly optimises the embedding space using normalised temperature-scaled cross entropy (NT-Xent) loss. We use this model to perform keyword spotting for radio broadcasts in Luganda and Bambara, the latter a severely under-resourced language. We compare our model to various existing AWE approaches, including those constructed from large pre-trained self-supervised models, a recurrent encoder which previously used the NT-Xent loss, and a DTW baseline. We demonstrate that the proposed contrastive transformer approach offers performance improvements over all considered existing approaches to very low-resource keyword spotting in both languages. |
| title | Low-resource keyword spotting using contrastively trained transformer acoustic word embeddings |
| topic | Audio and Speech Processing |
| url | https://arxiv.org/abs/2506.17690 |