Salvato in:
| Autori principali: | , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2506.03606 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916778048749568 |
|---|---|
| author | Gogoi, Parismita Kalita, Sishir Lalhminghlui, Wendy Terhiija, Viyazonuo Tzudir, Moakala Sarmah, Priyankoo Prasanna, S. R. M. |
| author_facet | Gogoi, Parismita Kalita, Sishir Lalhminghlui, Wendy Terhiija, Viyazonuo Tzudir, Moakala Sarmah, Priyankoo Prasanna, S. R. M. |
| contents | This study explores the use of self-supervised learning (SSL) models for tone recognition in three low-resource languages from North Eastern India: Angami, Ao, and Mizo. We evaluate four Wav2vec2.0 base models that were pre-trained on both tonal and non-tonal languages. We analyze tone-wise performance across the layers for all three languages and compare the different models. Our results show that tone recognition works best for Mizo and worst for Angami. The middle layers of the SSL models are the most important for tone recognition, regardless of the pre-training language, i.e. tonal or non-tonal. We have also found that the tone inventory, tone types, and dialectal variations affect tone recognition. These findings provide useful insights into the strengths and weaknesses of SSL-based embeddings for tonal languages and highlight the potential for improving tone recognition in low-resource settings. The source code is available at GitHub 1 . |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_03606 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Tone recognition in low-resource languages of North-East India: peeling the layers of SSL-based speech models Gogoi, Parismita Kalita, Sishir Lalhminghlui, Wendy Terhiija, Viyazonuo Tzudir, Moakala Sarmah, Priyankoo Prasanna, S. R. M. Audio and Speech Processing Artificial Intelligence Computation and Language Signal Processing This study explores the use of self-supervised learning (SSL) models for tone recognition in three low-resource languages from North Eastern India: Angami, Ao, and Mizo. We evaluate four Wav2vec2.0 base models that were pre-trained on both tonal and non-tonal languages. We analyze tone-wise performance across the layers for all three languages and compare the different models. Our results show that tone recognition works best for Mizo and worst for Angami. The middle layers of the SSL models are the most important for tone recognition, regardless of the pre-training language, i.e. tonal or non-tonal. We have also found that the tone inventory, tone types, and dialectal variations affect tone recognition. These findings provide useful insights into the strengths and weaknesses of SSL-based embeddings for tonal languages and highlight the potential for improving tone recognition in low-resource settings. The source code is available at GitHub 1 . |
| title | Tone recognition in low-resource languages of North-East India: peeling the layers of SSL-based speech models |
| topic | Audio and Speech Processing Artificial Intelligence Computation and Language Signal Processing |
| url | https://arxiv.org/abs/2506.03606 |