Tools
Tools
Kyi Toe, Wutyi (2024) Endangered Red Panda Behaviours Classification: A Comparative Evaluation of Deep Learning Models. Masters thesis, Dublin, National College of Ireland.
Preview |
PDF (Master of Science)
Download (5MB) | Preview |
Preview |
PDF (Configuration Manual)
Download (2MB) | Preview |
Abstract
The significant advancements of machine learning contribute to analysing the complex patterns of data and providing insightful information to the various fields. Therefore, this research is motivated to utilise machine learning techniques for the classification of the red panda behaviours. The red panda was listed as an endangered species in 2015 due to the decline of their population up to 50% over the last 20 years. The conservation of red pandas is critical to maintaining the ecosystem balance, and understanding the behaviours of red pandas is important to engage in their conservation in wildlife and zoos. This research proposes state-of-the-art Convolutional Neural Network (CNN) models to contribute to the classification of red panda behaviours (eating, resting, and walking) using images and video footage. As the Red Panda dataset is not available publicly, the custom Red Panda dataset is created specifically for this research work. A self-trained CNN model is developed and named CNN-RedPanda; moreover, pre-trained CNN models – EfficientNetB0 and ResNet50 are implemented using the transfer learning technique. The research highlights that ResNet50 is the most accurate model with 99.27% while EfficientNetB0 is balanced in accuracy and computational efficiency. The self-trained CNN-RedPanda model achieves considerably good performance at 91.35% accuracy and can be regarded as the most lightweight model among three of them. This research work contributes to the conservation of endangered red pandas with the novel approach and optimal results of the deep learning models.
| Item Type: | Thesis (Masters) |
|---|---|
| Supervisors: | Name Email Haque, Rejwanul UNSPECIFIED |
| Subjects: | G Geography. Anthropology. Recreation > GE Environmental Sciences Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning |
| Divisions: | School of Computing > Master of Science in Artificial Intelligence |
| Depositing User: | Ciara O'Brien |
| Date Deposited: | 20 Jun 2025 08:39 |
| Last Modified: | 20 Jun 2025 08:39 |
| URI: | https://norma.ncirl.ie/id/eprint/7954 |
Actions (login required)
 |
View Item |