Forensic License Plate Recognition with Compression-Informed Transformers

TL;DR

This paper introduces a compression-aware Transformer model for forensic license plate recognition, significantly improving accuracy on degraded images in legal investigations.

Contribution

It presents a novel side-informed Transformer architecture that incorporates compression knowledge, enhancing recognition of highly degraded license plates.

Findings

Outperforms existing FLPR methods and standard models.

Achieves up to 8.9% accuracy improvement on severely degraded images.

Requires fewer parameters than comparable models.

Abstract

Forensic license plate recognition (FLPR) remains an open challenge in legal contexts such as criminal investigations, where unreadable license plates (LPs) need to be deciphered from highly compressed and/or low resolution footage, e.g., from surveillance cameras. In this work, we propose a side-informed Transformer architecture that embeds knowledge on the input compression level to improve recognition under strong compression. We show the effectiveness of Transformers for license plate recognition (LPR) on a low-quality real-world dataset. We also provide a synthetic dataset that includes strongly degraded, illegible LP images and analyze the impact of knowledge embedding on it. The network outperforms existing FLPR methods and standard state-of-the art image recognition models while requiring less parameters. For the severest degraded images, we can improve recognition by up to 8.9 percent points.