Reading in the Dark: Low-light Scene Text Recognition
Xuanshuo Fu, Lei Kang, Ernest Valveny, Dimosthenis Karatzas, Javier Vazquez-Corral · 2026
Read Reading in the Dark: Low-light Scene Text Recognition by Xuanshuo Fu, Lei Kang, Ernest Valveny, Dimosthenis Karatzas, Javier Vazquez-Corral (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-2604-23685.txt.
Description
Fu and colleagues tackle text recognition in the dark, where dim light and noise defeat ordinary reading systems. They release LSTR, a dataset of over 11,000 low-light images derived from well-lit collections, plus a small set of real night street scenes for testing only. Jointly training an enhancement module with a text reader beats either part alone, and the team asks how bright is bright enough for reliable reading.
Catalog record
| CREATOR | Xuanshuo Fu, Lei Kang, Ernest Valveny, Dimosthenis Karatzas, Javier Vazquez-Corral |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 34144 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.23685