Tools
Tools
Sharma, Pravin Harish (2024) A Deep Learning Approach for Chicken Disease Detection using Images of Droppings. Masters thesis, Dublin, National College of Ireland.
Preview |
PDF (Master of Science)
Download (1MB) | Preview |
Preview |
PDF (Configuration Manual)
Download (2MB) | Preview |
Abstract
This research focuses on developing a lightweight deep learning pipeline for poultry disease detection and classification models based on fecal images. The study makes a thorough review of existing literature and concludes that an ensemble of lightweight transfer learning models will help keep the model size low, while still maintaining high accuracy. An ensemble of three lightweight transfer learning models was created using MobileNetV3, NASNetMobile and EfficientNetV2B2. Various experiments were conducted to optimize the models, utilizing techniques such as early stopping and learning rate adjustments. The averaging ensemble model for image classification achieved a validation accuracy of 98.91%, outperforming previous works, while the YOLOv10-S model attained an 89.5% mAP for object detection. The research demonstrates the effectiveness of using lightweight models in achieving high accuracy with fewer parameters, making the solution suitable for mobile deployment. Future work includes exploring ensemble techniques like bagging and boosting for further performance improvements.
| Item Type: | Thesis (Masters) |
|---|---|
| Supervisors: | Name Email Haycock, Barry UNSPECIFIED |
| Subjects: | Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science S Agriculture > SF Animal culture T Technology > TR Photography |
| Divisions: | School of Computing > Master of Science in Data Analytics |
| Depositing User: | Ciara O'Brien |
| Date Deposited: | 25 Aug 2025 11:15 |
| Last Modified: | 25 Aug 2025 11:15 |
| URI: | https://norma.ncirl.ie/id/eprint/8627 |
Actions (login required)
 |
View Item |