Tools
Tools
Ghosh, Sayan Kumar (2024) Improving Public Safety: Advanced Machine Learning for Early Detection of Aggressive Street Dog Behaviours in Asian Urban Environments. Masters thesis, Dublin, National College of Ireland.
Preview |
PDF (Master of Science)
Download (975kB) | Preview |
Preview |
PDF (Configuration Manual)
Download (1MB) | Preview |
Abstract
This paper introduces a novel multi-model computer vision system for identifying and monitoring the aggression and following behaviours in street dogs of Asia’s urban areas. Employing different types of models 2D Convolutional Neural Network, Vision transformer-b16, EfficientNet B3, ResNext -50 and amongst all and sundry, the 2D CNN came out to be very effective, with validation accuracy rate of 98. 12%. This system uses YOLO V5(You Only Look once) for real-time and fast dog detection, and DeepSORT for real-time tracking. It is also a strong factor since the system is compatible with city cameras, making it possible to greatly improve protection in urban environments and perform constant observation accompanied by instant responses, which are essential in complicated urban settings. However, it is confronted with difficulties such as obstacles which it may encounter in passing through, fast motion which a particular object might be displaying, and fluctuations in light situations that sometimes affects its outcomes in terms of brightness and sharpness. Such challenges call for the future improvement, specifically how best to enhance its performance especially in low resolution and dynamic environment scenarios. Through the application of deep learning algorithms, this system quickly scans for attributions of aggression in a dog that could potentially lead to conflicts among people, thus fostering the maintenance of public order. Possible ethical concerns related to the use of this technology to address the problem of stray dogs as well as the impacts of this technology on the urban populations are also discussed. The conclusion of this research work is essential for enhancing interphase relations between humans and dogs in Asian cities. The author also presents the measures applicable for preventing the aggressive behaviour and promoting the cooperation among people and animals.
| Item Type: | Thesis (Masters) |
|---|---|
| Supervisors: | Name Email Menghwar, Teerath Kumar UNSPECIFIED |
| Subjects: | Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science R Medicine > RA Public aspects of medicine > RA0421 Public health. Hygiene. Preventive Medicine Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning |
| Divisions: | School of Computing > Master of Science in Data Analytics |
| Depositing User: | Ciara O'Brien |
| Date Deposited: | 18 Aug 2025 15:07 |
| Last Modified: | 18 Aug 2025 15:07 |
| URI: | https://norma.ncirl.ie/id/eprint/8568 |
Actions (login required)
 |
View Item |