← Back to Catalog
.edu Email Required
🇧🇷UFPR-VeSV
Recognition + Vehicle Attributes · Brazil · 2026
Images
24,945
Frame Type
Single
Chars
36
Classes (+CTC blank)
37
Annotation
JSON: plate, corners, make, model, color, type, infrared, rear_view
Description
Multi-attribute vehicle surveillance dataset. 24.9K images with color, model, make annotations alongside plate text. Enables vehicle-aware LP recognition.
Resolution: ~480-720px typical
License: Academic (.edu required)
Access
Citation
BibTeX
@article{lima2026toward_vesv,
title={Toward Unified Fine-Grained Vehicle Classification and Automatic License Plate Recognition},
author={Lima, GE and Nascimento, V and Santos, E and Nascimento Jr, E and Laroca, R and Menotti, D},
journal={JBCS},
volume={32},
number={1},
year={2026}
}