Breaking the Transcription Bottleneck: Fine-tuning ASR Models for Extremely Low-Resource Fieldwork Languages
Fuente:
arXiv
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912445870637056 |
|---|---|
| author | Liang, Siyu Levow, Gina-Anne |
| author_facet | Liang, Siyu Levow, Gina-Anne |
| contents | Automatic Speech Recognition (ASR) has reached impressive accuracy for high-resource languages, yet its utility in linguistic fieldwork remains limited. Recordings collected in fieldwork contexts present unique challenges, including spontaneous speech, environmental noise, and severely constrained datasets from under-documented languages. In this paper, we benchmark the performance of two fine-tuned multilingual ASR models, MMS and XLS-R, on five typologically diverse low-resource languages with control of training data duration. Our findings show that MMS is best suited when extremely small amounts of training data are available, whereas XLS-R shows parity performance once training data exceed one hour. We provide linguistically grounded analysis for further provide insights towards practical guidelines for field linguists, highlighting reproducible ASR adaptation approaches to mitigate the transcription bottleneck in language documentation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_17459 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Breaking the Transcription Bottleneck: Fine-tuning ASR Models for Extremely Low-Resource Fieldwork Languages Liang, Siyu Levow, Gina-Anne Computation and Language Sound Audio and Speech Processing Automatic Speech Recognition (ASR) has reached impressive accuracy for high-resource languages, yet its utility in linguistic fieldwork remains limited. Recordings collected in fieldwork contexts present unique challenges, including spontaneous speech, environmental noise, and severely constrained datasets from under-documented languages. In this paper, we benchmark the performance of two fine-tuned multilingual ASR models, MMS and XLS-R, on five typologically diverse low-resource languages with control of training data duration. Our findings show that MMS is best suited when extremely small amounts of training data are available, whereas XLS-R shows parity performance once training data exceed one hour. We provide linguistically grounded analysis for further provide insights towards practical guidelines for field linguists, highlighting reproducible ASR adaptation approaches to mitigate the transcription bottleneck in language documentation. |
| title | Breaking the Transcription Bottleneck: Fine-tuning ASR Models for Extremely Low-Resource Fieldwork Languages |
| topic | Computation and Language Sound Audio and Speech Processing |
| url | https://arxiv.org/abs/2506.17459 |