Boosting Norwegian Automatic Speech Recognition
Javier De La Rosa, Rolv-Arild Braaten, Per Kummervold, Freddy Wetjen
Abstract
In this paper, we present several baselines for automatic speech recognition (ASR) models for the two official written languages in Norway: Bokmål and Nynorsk. We compare the performance of models of varying sizes and pre-training approaches on multiple Norwegian speech datasets. Additionally, we measure the performance of these models against previous state-of-the-art ASR models, as well as on out-of-domain datasets. We improve the state of the art on the Norwegian Parliamentary Speech Corpus (NPSC) from a word error rate (WER) of 17.10% to 7.60%, with models achieving 5.81% for Bokmål and 11.54% for Nynorsk. We also discuss the challenges and potential solutions for further improving ASR models for Norwegian.- Anthology ID:
- 2023.nodalida-1.55
- Volume:
- Proceedings of the 24th Nordic Conference on Computational Linguistics (NoDaLiDa)
- Month:
- May
- Year:
- 2023
- Address:
- Tórshavn, Faroe Islands
- Editors:
- Tanel Alumäe, Mark Fishel
- Venue:
- NoDaLiDa
- SIG:
- Publisher:
- University of Tartu Library
- Note:
- Pages:
- 555–564
- Language:
- URL:
- https://aclanthology.org/2023.nodalida-1.55
- DOI:
- Cite (ACL):
- Javier De La Rosa, Rolv-Arild Braaten, Per Kummervold, and Freddy Wetjen. 2023. Boosting Norwegian Automatic Speech Recognition. In Proceedings of the 24th Nordic Conference on Computational Linguistics (NoDaLiDa), pages 555–564, Tórshavn, Faroe Islands. University of Tartu Library.
- Cite (Informal):
- Boosting Norwegian Automatic Speech Recognition (De La Rosa et al., NoDaLiDa 2023)
- PDF:
- https://preview.aclanthology.org/nschneid-patch-5/2023.nodalida-1.55.pdf