Breeze Taigi: Benchmarks and Models for Taiwanese Hokkien Speech Recognition and Synthesis
Yu-Siang Lan, Chia-Sheng Liu, Yi-Chang Chen, Po-Chun Hsu, Allyson Chiu, Shun-Wen Lin, Da-shan Shiu, Yuan-Fu Liao, [0.5em] National Yang Ming Chiao Tung University · 2026
Read Breeze Taigi: Benchmarks and Models for Taiwanese Hokkien Speech Recognition and Synthesis by Yu-Siang Lan, Chia-Sheng Liu, Yi-Chang Chen, Po-Chun Hsu, Allyson Chiu, Shun-Wen Lin, Da-shan Shiu, Yuan-Fu Liao, [0.5em] National Yang Ming Chiao Tung University (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-2603-19259.txt.
Description
Lan and co-authors build open tools for Taiwanese Hokkien speech tech. Their Breeze Taigi sets a shared benchmark for speech recognition and synthesis in Taigi. It uses 30 paired Mandarin and Taigi clips from public service notices, with cleaned transcripts. Character error rate is set as the shared score, with fixed text cleanup rules. A Whisper model tuned on about 10,000 hours of synthetic Taigi speech leads the test. It posts a 30.13 percent error rate, ahead of prior systems.
Catalog record
| CREATOR | Yu-Siang Lan, Chia-Sheng Liu, Yi-Chang Chen, Po-Chun Hsu, Allyson Chiu, Shun-Wen Lin, Da-shan Shiu, Yuan-Fu Liao, [0.5em] National Yang Ming Chiao Tung University |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 34574 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.19259