Please use this identifier to cite or link to this item: https://hdl.handle.net/20.500.13087/2230
Title: Accurate CNN-based pupil segmentation with an ellipse fit error regularization term
Authors: Akınlar, Cüneyt
Küçükkartal, Hatice Kübra
Topal, Cihan
Keywords: Pupil segmentation
Convolutional Neural Networks (CNN)
UNet
Loss function
Regularization term
Eye-Tracking
Computational Approach
Gaze
Movements
Algorithm
Edge
Issue Date: 2022
Publisher: Pergamon-Elsevier Science Ltd
Abstract: Semantic segmentation of images by Fully Convolutional Neural Networks (FCN) has gained increased attention in recent years as FCNs greatly outperform traditional segmentation algorithms. In this paper we propose using Ellipse Fit Error as a shape prior regularization term that can be added to a pixel-wise loss function, e.g., binary cross entropy, to train a CNN for pupil segmentation. We evaluate the performance of the proposed method by training a lightweight UNet architecture, and use three widely used real-world datasets for pupil center estimation, i.e., ExCuSe, ElSe, and Labeled Pupils in the Wild (LPW), containing a total of similar to 230.000 images for performance evaluation. Experimental results show that the proposed method gives the best-known pupil detection rates for all datasets.
URI: http://doi.org/10.1016/j.eswa.2021.116004
https://hdl.handle.net/20.500.13087/2230
ISSN: 0957-4174
1873-6793
Appears in Collections:Bilgisayar Mühendisliği Bölümü Koleksiyonu
Scopus İndeksli Yayınlar Koleksiyonu
WoS İndeksli Yayınlar Koleksiyonu

Show full item record

CORE Recommender

SCOPUSTM   
Citations

1
checked on Dec 28, 2022

Page view(s)

114
checked on Oct 3, 2022

Google ScholarTM

Check

Altmetric


Items in GCRIS Repository are protected by copyright, with all rights reserved, unless otherwise indicated.