2020/10/22 by Renu Sharma, Arun Ross, Sharma, Renu +1 · 2 citations
Computer Science · #Biometric Identification and Security #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Machine Learning (cs.LG)
paper · pdf · doi:10.48550/arxiv.2011.10655
openalex publication_date 2020/10/22 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
In this paper, we propose the use of Optical Coherence Tomography (OCT)\nimaging for the problem of iris presentation attack (PA) detection. We assess\nits viability by comparing its performance with respect to traditional iris\nimaging modalities, viz., near-infrared (NIR) and visible spectrum. OCT imaging\nprovides a cross-sectional view of an eye, whereas traditional imaging provides\n2D iris textural information. PA detection is performed using three\nstate-of-the-art deep architectures (VGG19, ResNet50 and DenseNet121) to\ndifferentiate between bonafide and PA samples for each of the three imaging\nmodalities. Experiments are performed on a dataset of 2,169 bonafide, 177 Van\nDyke eyes and 360 cosmetic contact images acquired using all three imaging\nmodalities under intra-attack (known PAs) and cross-attack (unknown PAs)\nscenarios. We observe promising results demonstrating OCT as a viable solution\nfor iris presentation attack detection.\n