A comprehensive comparison analysis of scholarly investigations on deep neural network for Human Iris detection

Deb, Debmalya (2023) A comprehensive comparison analysis of scholarly investigations on deep neural network for Human Iris detection. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (8MB) | Preview
Preview	PDF (Configuration Manual) Download (7MB) | Preview

Abstract

The research in Iris and eye gaze detection was initiated long ago and since then the research community has been trying to solve many critical case studies for example cheating on online examinations, digital biometrics, driver’s eye fatigue detection during driving and many more which additionally has many business impacts. This paper is a comparison analysis of the performance and implementation of a high computing transfer learning model, Residual Network(ResNet) with a comparatively less compute unit consumption model, Convolution Neural Network(CNN). Images were captured from open-sourced videos and along with that, the real-time images were also been taken with the help of a built-in webcam. The taken images were annotated and stored in JSON documents and soon after they were augmented to get multiple features and scenarios. The augmented images were fed through the deep-learning models for evaluation and detection analysis. In the evaluation metrics, this study employed Mean Square Error(MSE) as validation and test loss, and some additional metrics like Root Mean Square Error(RMSE) and Mean Absolute Error(MAE). Finally, the project concluded that the high-computing ResNet model outperformed the less-computing CNN model regarding evaluation metrics and Iris detection analysis.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Anant, Aaloka UNSPECIFIED
Uncontrolled Keywords:	Iris detection; ResNet mode; CNN model; Key point Annotations; Data Augmentation
Subjects:	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 Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	07 May 2025 14:07
Last Modified:	07 May 2025 14:07
URI:	https://norma.ncirl.ie/id/eprint/7506

Actions (login required)

View Item