License Plate Images Generation with Diffusion Models

Mariia Shpir, Nadiya Shvai, Amir Nakib · 2025

Read License Plate Images Generation with Diffusion Models by Mariia Shpir, Nadiya Shvai, Amir Nakib (2025) 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-2501-03374.txt.

Description

Mariia Shpir, Nadiya Shvai, and Amir Nakib generate fake license plate images with diffusion models in a 2025 paper. Plate recognition research is starved of data because privacy law limits sharing real plates. Synthetic plates could fill the gap if they look real enough.

A diffusion model trained on a Ukrainian plate collection produced a thousand test images. The team sorted the outputs by hand to study success rates and failure types. Character mix and error patterns were charted in detail. The model proved able to make convincing plates at a useful rate. The authors also released a public set of ten thousand synthetic plates. Such data can train recognition systems without exposing real drivers.

Catalog record

CREATORMariia Shpir, Nadiya Shvai, Amir Nakib
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE32 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-03374

Browse the full library · About the Tower of Babel