Authors: Md. Shahriar Rahman Anuvab, Mishkat Sultana, Md. Atif Hossain, Shashwata Das, Suvarthi Chowdhury, Rafeed Rahman, Dibyo Fabian Dofadar, Shahriar Rahman Rana

Published on: April 07, 2024

Impact Score: 7.8

Arxiv code: Arxiv:2404.05049

Summary

• What is new: PlateSegFL introduces a U-Net-based segmentation combined with Federated Learning for improved performance in license plate recognition.
• Why this is important: Existing ALPR systems use one-shot learners or pre-trained models limited by geometric bounding boxes and face network and complexity issues with continuous video streams.
• What the research proposes: PlateSegFL uses U-Net for multi-class image segmentation and Federated Learning to reduce data transfer needs and protect privacy.
• Results: Achieved around 95% F1 score, indicating high accuracy and efficiency.

Technical Details

Technological frameworks used: Federated Learning, U-Net-based segmentation

Models used: nan

Data used: nan

Potential Impact

Transportation, vehicle communication systems, security companies, traffic management organizations.

Want to implement this idea in a business?

We have generated a startup concept here: SafeRoute.