Article
Computer Science, Information Systems
Esteban Villalobos, Domingo Mery, Kevin Bowyer
Summary: Facial recognition accuracy varies among different demographic groups. Using demographic-specific thresholds to mitigate this issue may be controversial and impractical. Predicted non-sensitive soft-biometric attributes can achieve a similar mitigation effect.
Article
Computer Science, Information Systems
Gayathri R. Nayar, Tony Thomas
Summary: Palm vein biometrics is a potential authentication technique that shows high accuracy and anti-spoofing capability. However, the large palm size restricts its applicability. This paper proposes using partial palm vein patterns for authentication and demonstrates the feasibility of using only a small portion of the palm vein pattern. The proposed authentication framework achieves good results on various datasets.
JOURNAL OF INFORMATION SECURITY AND APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Sangeeta Biswas, Johan Rohdin, Andrii Kavetskyi, Gabriel Saraiva, Angkan Biswas, Martin Drahansky
Summary: The research investigates the bilateral similarity of central retinal blood vessels in the left and right retinas for biometric identification. The performance of human- and deep neural network-based bilateral verification is evaluated and analyzed through experiments on publicly available datasets.
Article
Medicine, General & Internal
Carol Ren Lin, Abduqodir Toychiev, Reynolds Kwame Ablordeppey, Miduturu Srinivas, Alexandra Benavente-Perez
Summary: The integrity of the aged retinal microvasculature is affected by myopia-induced ocular growth, leading to changes in the number of string vessels and vessel branchpoints. These changes suggest an adaptive reorganization of the retinal microvascular structure during aging and myopia development.
FRONTIERS IN MEDICINE
(2023)
Article
Mathematics
Muhammad Arsalan, Adnan Haider, Ja Hyung Koo, Kang Ryoung Park
Summary: Retinal blood vessels are important biomarkers for ophthalmic diseases, but the manual analysis of vascular changes is a tedious task. This study proposes a new network model called VSUL-Net, which accurately extracts retinal vasculature from images without the need for complex preprocessing.
Article
Computer Science, Artificial Intelligence
He Liu, Tao Wang, Yidong Li, Congyan Lang, Yi Jin, Haibin Ling
Summary: This paper proposes a joint graph learning and matching network called GLAM to explore reliable graph structures for boosting graph matching. The experimental results show that GLAM outperforms previous state-of-the-art methods in visual matching tasks and remarkably enhances deep graph matching methods.
PATTERN RECOGNITION
(2023)
Article
Computer Science, Interdisciplinary Applications
Mohsin Challoob, Yongsheng Gao, Andrew Busch, Mohammad Nikzad
Summary: This paper introduces a new separable paravector orientation tensor that addresses the challenges of robust detection of retinal vessels. The proposed method characterizes vessel enhancement using a nonlinear scale representation, invariant to changes in contrast and lighting, responsive to symmetric patterns, and fitted with elliptical cross-sections. By projecting vessels as a 3D paravector valued function and incorporating eigensystem analysis to fit constraint elliptical volumes, the proposed method achieves excellent performance in detecting retinal vessels.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2023)
Article
Ophthalmology
Noam D. Rudnick, Filippos Vingopoulos, Jay C. Wang, Itika Garg, Ying Cui, Ying Zhu, Rongrong Le, Raviv Katz, Yifan Lu, Nimesh A. Patel, John B. Miller
Summary: This study used widefield swept-source optical coherence tomography angiography (WF SS OCTA) to characterize the morphology, location, and functional significance of macular and extramacular collateral vessels (CVs) in patients with a history of branch retinal vein occlusion (BRVO). The study found that a large fraction of CVs that form after BRVO are extramacular, and the morphology of CVs varies as a function of retinal location.
BRITISH JOURNAL OF OPHTHALMOLOGY
(2022)
Article
Geography, Physical
Liang Shen, Zhou Xu, Jiahua Zhu, Xiaotao Huang, Tian Jin
Summary: In this paper, we propose a novel probabilistic local verification method to reject false feature matches and establish reliable feature correspondence. The proposed method outperforms other existing methods in terms of success rate and F-measure on multiple public test datasets.
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING
(2022)
Review
Engineering, Electrical & Electronic
Borui Hou, Huijie Zhang, Ruqiang Yan
Summary: This article provides an overview of finger-vein-based biometric methods, including image acquisition, image preprocessing, feature extraction, and matching. It introduces template-based, representation-based, and learning-based methods, as well as deep-learning-based methods. The effectiveness of these methods is verified through case studies and the challenges and prospects of finger-vein recognition research, particularly 3-D finger-vein recognition research, are discussed to guide future research direction in this field.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2022)
Article
Engineering, Electrical & Electronic
Jothibasu Marappan, Karthik Murugesan, Malar Elangeeran, Usha Subramanian
Summary: Retinal biometric trait is highly secure for personal identification. Preprocessing technique is necessary for poor contrast retinal images, especially for vessel segmentation. Proposed method based on RCMs shape descriptors shows high accuracy and efficiency in blood vessel segmentation.
JOURNAL OF ELECTRONIC IMAGING
(2023)
Article
Computer Science, Theory & Methods
Danilo Avola, Marco Cascio, Luigi Cinque, Alessio Fagioli, Chiara Petrioli
Summary: Person re-identification is a challenging task that aims to recognize a person across different cameras using Wi-Fi signals. This paper proposes a method that extracts the amplitude and phase of Wi-Fi signals to achieve person re-identification. Experimental results demonstrate that Wi-Fi signal variations differ between individuals and can be used for Re-ID.
IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY
(2022)
Article
Computer Science, Artificial Intelligence
Jiayi Ma, Xingyu Jiang, Aoxiang Fan, Junjun Jiang, Junchi Yan
Summary: Image matching is a fundamental task in various visual applications, and with the development of deep learning techniques, there has been an increasing number of methods proposed in this field. However, the challenge remains in choosing the suitable method for specific applications and designing image matching methods with superior performance. This comprehensive review and analysis provide insights into classical and latest techniques, and offer prospects for future development in image matching technologies.
INTERNATIONAL JOURNAL OF COMPUTER VISION
(2021)
Article
Chemistry, Multidisciplinary
Roziana Ramli, Khairunnisa Hasikin, Mohd Yamani Idna Idris, Noor Khairiah A. Karim, Ainuddin Wahid Abdul Wahab
Summary: This study introduces a feature-based retinal fundus image registration technique that combines retinal vessels and noise characteristics to increase accuracy. The registration method CURVE-SIFT outperformed existing methods significantly, achieving a higher success rate in image pairing.
APPLIED SCIENCES-BASEL
(2021)
Article
Biochemistry & Molecular Biology
Maria Hernandez, Jorge Gonzalez-Zamora, Sergio Recalde, Maite Moreno-Orduna, Valentina Bilbao-Malave, Manuel Saenz de Viteri, Manuel F. Landecho, Patricia Fernandez-Robredo, Alfredo Garcia-Layana
Summary: This study evaluated vascular and histological alterations in COVID-19 retinas, finding a decrease in vessel number and an increase in vessel tortuosity. Immunofluorescence analysis showed elevated expression of ACE2 and Iba1 in COVID-19 retinas, indicating that SARS-CoV-2 affects endothelial cells and immune activation in the retina.