Biometrics and Identity Science

Accepted Papers

Localization using Multi-Focal Spatial Attention for Masked Face Recognition. Yooshin Cho (KAIST), Hanbyel Cho (KAIST), Hyeong Gwon Hong (KAIST), Jaesung Ahn (KAIST), Dongmin Cho (Alchera inc.), JungWoo Chang (Alchera), Junmo Kim (KAIST)
Since the beginning of world-wide COVID-19 pandemic, facial masks have been recommended to limit the spread of the disease. However, these masks hide certain facial attributes. Hence, it has become difficult for existing face recognition systems to perform identity verification on masked faces. In this context, it is necessary to develop masked Face Recognition (MFR) for contactless biometric recognition systems. Thus, in this paper, we propose Complementary Attention Learning and Multi-Focal Spatial Attention that precisely removes masked region by training complementary spatial attention to focus on two distinct regions: masked regions and backgrounds. In our method, standard spatial attention and networks focus on unmasked regions, and extract mask-invariant features while minimizing the loss of the conventional Face Recognition (FR) performance. For conventional FR, we evaluate the performance on the IJB-C, Age-DB, CALFW, and CPLFW datasets. We evaluate the MFR performance on the ICCV2021-MFR/Insightface track, and demonstrate the improved performance on the both MFR and FR datasets. Additionally, we empirically verify that spatial attention of proposed method is more precisely activated in unmasked regions.

Improving Performance of Facial Biometrics With Quality-Driven Dataset Filtering. Iurii Medvedev (University of Coimbra); Nuno Goncalves (University of Coimbra).
Advancements in deep learning techniques and availbility of large scale face datasets led to significant performance gains in face recognition in recent years. Modern face recognition algorithms are trained on large-scale in-the-wild face datasets. At the same time, many facial biometric applications rely on controlled image acquisition and enrolment procedures (for instance, document security applications). That is why such face recognition approaches can demonstrate the deficiency of the performance in the target scenario (ICAO-compliant images). However, modern approaches for face image quality estimation may help to mitigate that problem. In this work, we introduce a strategy for filtering training datasets by quality metrics and demonstrate that it can lead to performance improvements in biometric applications that rely on face image modality. We filter the main academic datasets using the proposed filtering strategy and present performance metrics.

Assessing the Efficacy of a Self-Stigma Reduction Mental Health Program with Mobile Biometrics: Work-in-Progress. Nele Loecher (University of South Florida); Sayde L King (University of South Florida); Joseph Cabo (University of South Florida); Tempestt Neal (USF); Kristin Kosyluk (University of South Florida)
One of the strongest predictors of success in post-secondary education is student engagement. Unfortunately, people with psychiatric disabilities are less engaged in their campus communities. This work-in-progress paper details the disclosure-based self-stigma reduction program, Up To Me, which is developed to increase inclusion and engagement of people with mental illness on college campuses by teaching strategies to weigh costs and benefits of disclosing one's mental illness. Further, we elaborate on the program's evaluation mechanisms, which involve both self-reported and passively recorded smartphone sensor data. The latter reflects a unique merging of behavioral and computer sciences that serves to facilitate behavioral modeling using artificial intelligence as an objective measure of Up to Me outcomes. Similar to data collection for some activity and biometric recognition applications, we employ a publicly available and free-to-use smartphone sensor reading app to correlate self-reported well-being with Up to Me participant behaviors. We anticipate that the behavioral data gathered via smartphones will substantiate self-report data on Up to Me outcomes.

Hierarchically Organized Computer Vision in Support of Multi-Faceted Search for Missing Persons. Arturo Miguel Russell Bernal (University of Notre Dame); Jane Cleland-Huang (University of Notre Dame).
Missing person searches are typically initiated with a description of a person that typically includes their age, race, clothing, and gender, possibly supported by a photo. Unmanned Aerial Vehicles (UAVs) imbued with Computer Vision (CV) capabilities, can be deployed to quickly search an area to find the missing person; however, the search task is far more difficult when a crowd of people is present, and only the missing person must be identified. Deploying CV models and supporting code to effectively perform this task on the limited resources of a UAV is challenging, and can lead to considerable performance degradation. We therefore propose a model that hierarchically combines multiple CV models, exploits computing capabilities deployed on and off the UAV, and engages humans in the search using practices of human-machine teaming. Finally, we apply the hierarchical CV approach to show how a person's image, extracted from our new aerial dataset of occluded people can be matched to a description of the person provided in the dataset's metadata.

Workshop Schedule (Local Waikoloa, Hawaii, USA Time (HST))

Paper Submissions

Paper submissions for InterID 2023 will be handled through the CMT system. Please submit your article by the September 12, 2022 deadline. According to FG guidelines, papers longer than six pages will be subject to a page fee (100 USD per page) for the extra pages (two max). Submissions have to be in pdf format and are limited to a 10MB file size. Papers that use different formatting from the FG 2023 Latex or Word templates have more than eight pages, or explicitly reveal the authors’ identity will be automatically removed from the reviewing process. Please familiarize yourself with ALL FG submission guidelines prior to submitting to InterID 2023.

Work-in-Progress Papers

We encourage authors to submit Work-in-Progress (WiP) papers based on their recent viewpoints, new discoveries, new application challenges, visionary ideas, and early-stage design and development in disciplines that are in line with the areas of interest of InterID, yet not quite ready for a full technical paper submission. WiP papers are geared towards early career researchers, but also caters to seasoned scholars for the presentation of unfinished projects. WiP submissions thus provide a unique opportunity for exchanging brave new ideas, for receiving feedback on projects currently in progress and for fostering collaborations. WiP papers enable authors to share late-breaking results with the research community. The emphasis of WIP papers is on the novelty of the work, not completeness. Thus, WiP papers should describe the underlying ideas and motivations for the work in sufficient detail to allow the reviewers to assess the potential contribution of the work. The paper should also state explicitly how the work is “in progress”; explain the future directions and challenges for the project, and state how the work reported in this paper contributes to that bigger research agenda.

Position Papers

Position papers should present an arguable opinion about an issue relevant to InterID topics of interest. The argument should be valid and defensible, and supported by well-founded knowledge. It is important to support your argument with evidence to ensure the validity of your claims, as well as to refute the counterclaims to show that you are well informed about both sides.

Technical (Full) Papers

Technical papers must deliver a complete contribution, that is, work that is defined clearly and executed completely. Appropriate background research, experimental comparisons, and research hypotheses must be provided. Results should be thoroughly explained, and it should be clear how the discussed work contributes toward advancing science.

Collaboration Statement

All paper submissions are required to have a small section (no more than one paragraph) dedicated to explaining the interdisciplinary nature of the work (e.g., how the combination of disciplines influenced the research, collaborative mechanisms, sources of multimodal data across disciplines, etc.). This section is meant to highlight promising collaborative mechanisms that could be adopted by other researchers, and to also bring attention to current collaborative teams. Here's an example:

This work is a collaborative effort between the Food and Nature Lab (FNL) at the University of Kind People and the Chemical Engineering Lab (CEL) at the Institute of Determination. The FNL and CEL has collaborated on data collection to understand the evolution of lava in three different countries over the past two years. This manuscript details the application of chemical engineering to understand the destructive properties of heat transfer from lava particles potentially impacting neighboring communities.

***For the initial anonymous paper submission, please be sure not to reveal who the authors are:***

This work is a collaborative effort between the Lab 1 at the Institution 1 and the Lab 2 at the Institution 2. The Lab 1 and Lab 2 has collaborated on data collection to understand the evolution of lava in three different countries over the past two years. This manuscript details the application of chemical engineering to understand the destructive properties of heat transfer from lava particles potentially impacting neighboring communities.

We realize that your work may not involve an interdisciplinary team, although it still aligns with a topic of interest to InterID. In this case, the collaboration statement should briefly detail how your paper could influence other fields and/or have impact beyond your direct area of expertise. Here's an example:

This work leverages the sensing capabilities of smart devices to better understand work-life balance. The broader impacts of this work could impact employee engagement strategies, improve the mental health within the workforce, and lead to the development of practical policies for employee leave.

Keynote Speaker

Vitomir Štruc, Ph.D.

Dr. Vitomir Štruc is a Full Professor at the University of Ljubljana, Slovenia. He received his doctoral degree from the Faculty of Electrical Engineering in Ljubljana in 2010. Vitomir's research interests include problems related to biometrics, computer vision, image processing, pattern recognition and machine learning. He (co-)authored more than 150 research papers for leading international peer reviewed journals and conferences in these and related areas. He served in different capacities on the organizing committees of several top-tier vision conferences, including IEEE Face and Gesture, ICB, WACV and IJCB. Vitomir is a Senior Area Editor for the IEEE Transactions on Information Forensics and Security, a Subject Editor for Elsevier’s Signal Processing and an Associate Editor for Pattern Recognition, and IET Biometrics. He served as an Area Chair for WACV 2018, 2019, 2020, ICPR 2018, Eusipco 2019 and FG 2020, 2021 and as the Program Chair for IJCB 2020 and IWBF 2022. Currently he acts as a Program Co-Chair for IWBF 2023, workshop co-chair for WACV 2022, a Workshop and tutorial Co-Chair for FG 2023 and a General Co-Chair for IJCB 2023. Dr. Struc is a Senior member of the IEEE, a member of IAPR, EURASIP, Slovenia’s national contact point for the European Association for Biometrics (EAB) and the former president and current executive committee member of the Slovenian Pattern Recognition Society, the Slovenian branch of IAPR. Vitomir is also the current VP Technical Activities for the IEEE Biometrics Council.

Keynote Address: Generative Models in Computer Vision: From Fashion and Beauty to Biometrics

With the developments in the field of generative modeling and with the appearance of powerful model architectures, such as Generative Adversarial Networks (GAN), a wide range of new techniques and inventive algorithms has emerged recently to solve diverse computer vision problems, including many problems in the area of biometrics. In this talk, I will first provide a short overview of recent advances in generative modeling and describe some of our research efforts that focus explicitly on generative models. Next, I will present examples of our recent work utilizing generative models and talk about: (i) face image editing with our GAN inversion based MaskFaceGAN technique that allows for photo realistic image manipulation and explicitly addresses the problem of attribute entanglement seen with many latent-space based editing solutions, (ii) virtual try-on with our context-driven C-VTON model and GlassesGAN approach, and (ii) bimodal ocular image generation (and annotation) with our Dual-Branch StlyGAN2 (DB-StyleGAN) model. Finally, I will elaborate on some of the existing challenges with generative models and highlight future research directions.

Panelists

Domenic Forte, Steven A. Yatauro Faculty Fellow, Electrical and Computer Engineering Department, University of Florida

Dr. Domenic Forte is an Associate Professor with the Electrical and Computer Engineering (ECE) Department at University of Florida (UF). He received a B.S. degree in Electrical Engineering from Manhattan College in 2006, and the M.S. and Ph.D. degrees in Electrical Engineering from the University of Maryland, College Park, in 2010 and 2013, respectively. His research interests include hardware security primitive design and evaluation, hardware obfuscation, hardware Trojan detection and prevention, electronics supply chain security, security-aware design automation tools, reverse engineering and anti-reverse engineering, and biometrics. Dr. Forte is a recipient of the Presidential Early Career Award for Scientists and Engineers (PECASE) in 2019, the Early Career Award for Scientists and Engineers (ECASE) by Army Research Office (ARO) in 2019, the NSF Faculty Early Career Development Program (CAREER) Award in 2017, the ARO Young Investigator Award in 2016, the Northrop Grumman Fellowship in 2012, and the George Corcoran Memorial Outstanding Teaching Award in 2008. His research has been recognized through best paper awards and nominations from IJCB 2017, ISTFA 2017, HOST 2016, HOST 2015, DAC 2012, and AHS 2011.

Giti Javidi, Professor, Muma College of Business, University of South Florida

Giti Javidi is a professor of Cybersecurity at the School of Information Systems and Management, Muma College of Business. She also serves as the academic director of the Information Assurance and Cybersecurity Management undergraduate program. She joined USF in 2016, having previously served at Virginia State University as a professor of Computer Science at the College of Engineering. Professor Javidi’s research interests cut across several areas including design solutions for secure smart devices, cybersecurity management and analytics, and data analytics applications in business. She directs the Applied Research Collaborative (ARC) lab on the Sarasota-Manatee campus where crossdisciplinarity faculty, students and industry partners come together to carry out applied research and on-demand training. Her research has been published in leading computing journals, and she has been awarded a number of prestigious grants from external agencies including NSF, NASA, Google and Microsoft. As a long-time advocate for diversity and inclusion in the technology workforce, she has spearheaded a number of national and international projects in this domain throughout her academic career. Prof. Javidi has received several prestigious awards in recognition of her influential role in leading national efforts for inclusive STEM education and workforce. These awards include the 2015 Virginia Fearless Leaders, 2017 and 2020 Sarasota Women of Influence and 2018 USF Women in Leadership and Philanthropy. She earned a PhD and master’s degree from the University of South Florida and a bachelor’s degree in computer science from the University of Central Oklahoma.

Kristin Kosyluk, Department of Mental Health Law and Policy, University of South Florida

Kristin Kosyluk is an assistant professor in the Department of Mental Health Law & Policy and faculty affiliate with the Louis de la Parte Mental Health Institute. Dr. Kosyluk received her Ph.D. in Psychology from Illinois Institute of Technology’s Rehabilitation Counseling Education program in 2014. Dr. Kosyluk’s research agenda is defined by a focus on mental illness and psychiatric disability, with a special interest in social justice issues and stigma. Much of her work to date has investigated how stigma interferes with outcomes for this population. As a rehabilitation counselor, she recognizes the crucial role that vocation plays in the lives of individuals with disabilities, and has undertaken work in the area of employment and postsecondary education.

Yael Bensoussan, Assistant Professor, College of Medicine Otolaryngology, University of South Florida

Dr. Yael Bensoussan is an attending laryngologist and an Assistant Professor of Otolaryngology at the University of South Florida. She is a fellowship-trained laryngologist with advanced expertise in voice, swallowing, and upper airway evaluation and treatment. Prior to joining the University of South Florida, she completed her medical school at the University of Montreal, and her surgical residency training at the University of Toronto, Canada. She went on to complete a fellowship in Laryngology (voice, airway, and swallowing disorders) at the University of Southern California under the direction and mentorship of Dr. Michael Johns and Dr. Karla O’Dell. Dr. Bensoussan’s background also includes a degree in Speech Pathology and a prior career in music, which has led her to her passion for the voice and for laryngeal disorders. From a clinical standpoint, Dr. Bensoussan aims to provide expert care to patients with various laryngeal disorders and swallowing disorders including laryngeal cancer, vocal fold paralysis, benign laryngeal lesions, laryngotracheal stenosis, spasmodic dysphonia, muscle tension dysphonia, Zenkers’ diverticulum, and oropharyngeal dysphagia amongst others. She is known to provide compassionate and individualized care to all patients. She is also an avid proponent of multi-disciplinary care and has led multiple patient-education initiatives around tracheostomy care. While completing her clinical fellowship, she also obtained her Master of Science (MSc) in System Leadership and Innovations and the Institute of Health Policy, Evaluation and Management at the University of Toronto, Canada, where she worked on applications of Artificial Intelligence (AI) to the laryngology field. She is developing an expertise in the integration of AI into the field of voice research and is involved in many projects of healthcare digitization and data-driven care. Passionate about optimization of health systems efficiency and algorithms of care, she is building bridges with multiple partners to integrate technology to optimize the quality of care provided to her patients. Dr. Bensoussan has served on the Voice Committee of the American Laryngological Association (ALA), and was awarded the Young Faculty/ Practitioner Award of the ALA in 2021 for her research work on AI and gender-affirming voice care.

Simon Khan, Research Scientist, Air Force Research Laboratory

Dr. Simon Khan is a research scientist at Air Force Research Laboratory (AFRL), Rome, NY and has a PhD in Electrical and Computer Engineering program from Clarkson University under STEM+M fellowship from the Air Force. He holds a B.S in Electrical Engineering from Stonybrook University, NY, an M.S in Information System Management from National University, San Diego, CA and an M.S in Computer Engineering from Syracuse University, NY. He has been working for 15 years under different branches of the US government. He is the recipient of Brave Zulu Awards from the Department of Navy and multiple other awards for successful design and development of network and C4I systems. His current research interests are about behavioral biometrics specially mouse dynamics and GUI events, explainable artificial intelligence (XAI) and efficient transfer learning.