Enhancing the Theft Detection of Vehicle and Emergency Alert Using Novel ResNet -50 over K-Nearest Neighbors Classifier

Girish Subash, S. Kalaiarasi

2023

Abstract

This study seeks to bolster the security of vehicles in public car parks by advancing automated vehicle theft detection using two unique machine learning methods. We've opted for the Novel Resnet-50 and K-Nearest Neighbours classifiers. Their performance was assessed to gauge their proficiency in vehicle theft detection. With 80% of the dataset used for training and the remaining 20% for testing, using a sample size of ten, the performance of the Novel Resnet-50, which utilises facial recognition to enhance vehicle safety, was pitted against K-Nearest Neighbours. The former posted an impressive 97% accuracy, showcasing its prowess in spotting unauthorised users, while the latter recorded a 94% accuracy, with a significance level of 0.005 (p<0.05). Evidently, the Novel Resnet-50's integration of facial recognition offers a promising avenue in vehicle security enhancement compared to K-Nearest Neighbours.

Download


Paper Citation


in Harvard Style

Subash G. and Kalaiarasi S. (2023). Enhancing the Theft Detection of Vehicle and Emergency Alert Using Novel ResNet -50 over K-Nearest Neighbors Classifier. In Proceedings of the 1st International Conference on Artificial Intelligence for Internet of Things: Accelerating Innovation in Industry and Consumer Electronics - Volume 1: AI4IoT; ISBN 978-989-758-661-3, SciTePress, pages 415-422. DOI: 10.5220/0012505300003739


in Bibtex Style

@conference{ai4iot23,
author={Girish Subash and S. Kalaiarasi},
title={Enhancing the Theft Detection of Vehicle and Emergency Alert Using Novel ResNet -50 over K-Nearest Neighbors Classifier},
booktitle={Proceedings of the 1st International Conference on Artificial Intelligence for Internet of Things: Accelerating Innovation in Industry and Consumer Electronics - Volume 1: AI4IoT},
year={2023},
pages={415-422},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012505300003739},
isbn={978-989-758-661-3},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Artificial Intelligence for Internet of Things: Accelerating Innovation in Industry and Consumer Electronics - Volume 1: AI4IoT
TI - Enhancing the Theft Detection of Vehicle and Emergency Alert Using Novel ResNet -50 over K-Nearest Neighbors Classifier
SN - 978-989-758-661-3
AU - Subash G.
AU - Kalaiarasi S.
PY - 2023
SP - 415
EP - 422
DO - 10.5220/0012505300003739
PB - SciTePress