Smartphone-based Iris Recognition Technology
Naveenkumar Venkataswamy
PhD Candidate, Dept. of ECE
Clarkson University
Abstract: As smartphones play an increasingly important role in our daily lives, the need for stronger and more reliable security features is becoming essential. Iris recognition, a technology known for its high accuracy, has traditionally relied on near-infrared (NIR) imaging, which requires specialized cameras. But what if we could achieve the same level of security using the cameras that are already on our phones? This talk will introduce an innovative approach to iris recognition that leverages the visible light spectrum, making it possible to use regular smartphone cameras for this advanced biometric authentication. During the talk, we’ll explore how we developed a custom Android application that consistently captures high-quality iris images. The app is designed to automatically adjust focus and zoom to ensure the best possible image, using cutting-edge deep learning models like YOLOv3-tiny for detecting the eye and iris, and a streamlined version of the U-Net model, known as Ghost-Attention U-Net, to accurately segment the iris. The app also adheres to international standards to guarantee that the images meet the necessary quality for reliable recognition. In addition, we’ll discuss the creation of the CUVIRIS dataset, which includes a wide range of iris images captured using both NIR and visible light from a diverse group of 47 participants. This dataset serves as a valuable resource for testing and improving iris recognition technology. The talk will highlight how our approach has the potential to significantly advance mobile security by making high-quality iris recognition accessible to everyone with a smartphone.
Join us to learn how this research is bridging the gap between advanced biometric technology and everyday mobile applications and how it could lead to more secure and user-friendly smartphone experiences in the future.
Bio: Naveen received his B.S. degree in Electronics and Communication from VTU, KA, India. After gaining experience as an engineer at several companies in India, he is now pursuing a PhD in Electrical and Computer Engineering. His research is centered on developing methods to improve biometric recognition performance, with a special emphasis on iris and face recognition. He is passionate about creating portable biometric systems and integrating AI and deep learning into these technologies.
Thursday, September 5, 2024, 12:15-1:15 pm, CAMP 194
Join Link: https://clarkson.zoom.us/j/97763004044?pwd=fReadMi2o7OYVOIOgYm5yAuGGnbmdy.1
Contact for queries: Prof Masudul Imtiaz, mimtiaz@clarkson.edu
*Co-Sponsored by IEEE student branch and HKN
________________________________________________________________ Electrical and Computer Engineering l CLARKSON UNIVERSITY l Potsdam, New York 13699-5720