Zapisane w:
| Główni autorzy: | , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Wydane: |
2025
|
| Hasła przedmiotowe: | |
| Dostęp online: | https://arxiv.org/abs/2508.19721 |
| Etykiety: |
Dodaj etykietę
Nie ma etykietki, Dołącz pierwszą etykiete!
|
| _version_ | 1866909755351498752 |
|---|---|
| author | Carvalho, Carlos Teixeira, Francisco Botelho, Catarina Pompili, Anna Solera-Ureña, Rubén Paulo, Sérgio Julião, Mariana Rolland, Thomas Mendonça, John Pereira, Diogo Trancoso, Isabel Abad, Alberto |
| author_facet | Carvalho, Carlos Teixeira, Francisco Botelho, Catarina Pompili, Anna Solera-Ureña, Rubén Paulo, Sérgio Julião, Mariana Rolland, Thomas Mendonça, John Pereira, Diogo Trancoso, Isabel Abad, Alberto |
| contents | Existing resources for Automatic Speech Recognition in Portuguese are mostly focused on Brazilian Portuguese, leaving European Portuguese (EP) and other varieties under-explored. To bridge this gap, we introduce CAMÕES, the first open framework for EP and other Portuguese varieties. It consists of (1) a comprehensive evaluation benchmark, including 46h of EP test data spanning multiple domains; and (2) a collection of state-of-the-art models. For the latter, we consider multiple foundation models, evaluating their zero-shot and fine-tuned performances, as well as E-Branchformer models trained from scratch. A curated set of 425h of EP was used for both fine-tuning and training. Our results show comparable performance for EP between fine-tuned foundation models and the E-Branchformer. Furthermore, the best-performing models achieve relative improvements above 35% WER, compared to the strongest zero-shot foundation model, establishing a new state-of-the-art for EP and other varieties. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_19721 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | CAMÕES: A Comprehensive Automatic Speech Recognition Benchmark for European Portuguese Carvalho, Carlos Teixeira, Francisco Botelho, Catarina Pompili, Anna Solera-Ureña, Rubén Paulo, Sérgio Julião, Mariana Rolland, Thomas Mendonça, John Pereira, Diogo Trancoso, Isabel Abad, Alberto Computation and Language Audio and Speech Processing Existing resources for Automatic Speech Recognition in Portuguese are mostly focused on Brazilian Portuguese, leaving European Portuguese (EP) and other varieties under-explored. To bridge this gap, we introduce CAMÕES, the first open framework for EP and other Portuguese varieties. It consists of (1) a comprehensive evaluation benchmark, including 46h of EP test data spanning multiple domains; and (2) a collection of state-of-the-art models. For the latter, we consider multiple foundation models, evaluating their zero-shot and fine-tuned performances, as well as E-Branchformer models trained from scratch. A curated set of 425h of EP was used for both fine-tuning and training. Our results show comparable performance for EP between fine-tuned foundation models and the E-Branchformer. Furthermore, the best-performing models achieve relative improvements above 35% WER, compared to the strongest zero-shot foundation model, establishing a new state-of-the-art for EP and other varieties. |
| title | CAMÕES: A Comprehensive Automatic Speech Recognition Benchmark for European Portuguese |
| topic | Computation and Language Audio and Speech Processing |
| url | https://arxiv.org/abs/2508.19721 |