# An advanced combination of semi-supervised Normalizing Flow & Yolo (YoloNF) to detect and recognize vehicle license plates

arXiv: https://arxiv.org/abs/2207.10777 · HTML: https://mlsanity.com/paper/2207.10777/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 9 of its 31 references that are in the map.
- First posted on arXiv 21 Jul 2022.

## Builds on (most cited first, 9 of 9)

- [You Only Look Once: Unified, Real-Time Object Detection](https://mlsanity.com/paper/1506.02640/) (2015)
- [Fast R-CNN](https://mlsanity.com/paper/1504.08083/) (2015)
- [YOLOv3: An Incremental Improvement](https://mlsanity.com/paper/1804.02767/) (2018)
- [Density estimation using Real NVP](https://mlsanity.com/paper/1605.08803/) (2016)
- [Generative Adversarial Networks: An Overview](https://mlsanity.com/paper/1710.07035/) (2017)
- [Grad-CAM: Why did you say that?](https://mlsanity.com/paper/1611.07450/) (2016)
- [Bag of Freebies for Training Object Detection Neural Networks](https://mlsanity.com/paper/1902.04103/) (2019)
- [Segmentation-free Vehicle License Plate Recognition using ConvNet-RNN](https://mlsanity.com/paper/1701.06439/) (2017)
- [Open data for Moroccan license plates for OCR applications : data collection, labeling, and model construction](https://mlsanity.com/paper/2104.08244/) (2021)

## Abstract (arXiv)

Fully Automatic License Plate Recognition (ALPR) has been a frequent research topic due to several practical applications. However, many of the current solutions are still not robust enough in real situations, commonly depending on many constraints. This paper presents a robust and efficient ALPR system based on the state-of-the-art YOLO object detector and Normalizing flows. The model uses two new strategies. Firstly, a two-stage network using YOLO and a normalization flow-based model for normalization to detect Licenses Plates (LP) and recognize the LP with numbers and Arabic characters. Secondly, Multi-scale image transformations are implemented to provide a solution to the problem of the YOLO cropped LP detection including significant background noise. Furthermore, extensive experiments are led on a new dataset with realistic scenarios, we introduce a larger public annotated dataset collected from Moroccan plates. We demonstrate that our proposed model can learn on a small number of samples free of single or multiple characters. The dataset will also be made publicly available to encourage further studies and research on plate detection and recognition.