Comparative Analysis Of Various Machine Learning Models For Child Safety And Security System For Protecting Them From Child Trafficking And Assault

Authors

  • Anuja Jadhav , Nisarg Gandhewar

DOI:

https://doi.org/10.47750/pnr.2022.13.S09.152

Abstract

A novel algorithm is compared to a few different models that have been developed for the purpose of keeping children safe. The performance of the algorithm is evaluated by performing trials on a dataset and analysing the results of those trials. In this study, we conduct a comparison of the proposed method to three different types of neural networks: convolutional neural networks, recurrent neural networks, and long-term short-term memory. Neither of these types of comparisons has been done before. Photos that YOLOv4 deems to have questionable behaviour are marked as such, and the subsequent stage of the process, deep feature extraction, takes these pictures into consideration. The proposed model for the detection, alert, and classification of unknown behaviour produces classifiable findings through the use of the LCNN approach that was developed. The proposed HUADM-LCNN has a higher accuracy than CNN, RCNN, and LSTM, with a value of 0.87 being the benchmark for its performance. The findings of this study lead the researchers to the conclusion that the suggested activity identification and classification model, which includes the HUADM-LCNN implementation, provides performance that is superior to that provided by the standard classifier approaches. The proposed model for activity detection and categorization is therefore superior to the strategies that came before it in terms of its performance.

Downloads

Published

2022-11-17 — Updated on 2022-11-17

Versions

How to Cite

Anuja Jadhav , Nisarg Gandhewar. (2022). Comparative Analysis Of Various Machine Learning Models For Child Safety And Security System For Protecting Them From Child Trafficking And Assault. Journal of Pharmaceutical Negative Results, 1280–1287. https://doi.org/10.47750/pnr.2022.13.S09.152

Issue

Section

Articles