Sometin Beta Pass Notin (SBPN): Improving Multilingual ASR for Nigerian Languages via Knowledge Distillation
Sewade Ogun · 2026
Read Sometin Beta Pass Notin (SBPN): Improving Multilingual ASR for Nigerian Languages via Knowledge Distillation by Sewade Ogun (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-17710.txt.
Description
Sewade Ogun tackles speech recognition for Nigerian languages. Scarce data, shifting spelling, tone marks, accents, code-switching, and local names drag accuracy below English and French. A two-stage distillation method first learns from existing single-language models guided by language-specific word models. It then refines itself on machine-labeled data. The result, Sometin Beta Pass Notin, covers Yoruba, Hausa, Igbo, Nigerian Pidgin, and Nigerian English, cutting relative word error by 29% on average against single-language baselines.
Catalog record
| CREATOR | Sewade Ogun |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 51571 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.17710