Evaluation of the Detectron2 framework for Instance Segmentation of Multi-Component Meal Images

Adedeji, Ayodeji Michael (2024) Evaluation of the Detectron2 framework for Instance Segmentation of Multi-Component Meal Images. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (7MB) | Preview
Preview	PDF (Configuration Manual) Download (1MB) | Preview

Abstract

The task of instance segmentation of multi-component meal images presents a unique challenge due to the complex visual characteristics of certain food components, such as small areas, complex boundaries and the lack of visual distinction between food components. These challenges are further complicated by the diversity of food globally, with differences in preparation and presentation. This study uses a multi-stage segmentation approach with the Detectron2 framework on two multi-component meal image datasets, UECFOODPIXComplete and FoodSeg103, to assess the framework’s suitability at the task of instance segmentation for multicomponent meal images. Experimental results showed the framework achieved a mean average precision score of 36.4% for the UECFOODPIXComplete dataset and 20.9% for the FoodSeg103 dataset. This research highlights the framework’s potential, challenges and areas of improvement at the task of instance segmentation of multi-component meal images.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Trinh, Anh Duong UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence > Computer vision Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence > Computer vision H Social Sciences > HD Industries. Land use. Labor > Specific Industries > Food Industry
Divisions:	School of Computing > Master of Science in Artificial Intelligence
Depositing User:	Ciara O'Brien
Date Deposited:	19 Jun 2025 15:17
Last Modified:	19 Jun 2025 15:17
URI:	https://norma.ncirl.ie/id/eprint/7940

Actions (login required)

View Item