Research
Privacy-Preserving Digital Identity Platform
Identity checks are required everywhere - be it in a physical store for purchasing age-restricted items (e.g. alcohol) or while shopping online to obtain special discounts (e.g., military or student). Manual identity check mechanisms often result in high cost of operations, inadequate protection of consumer privacy, and significant potential of fraud. In this research, we design a digital identity platform that enables relying parties to perform automatic consumer identity checks using data supplied by identity providers (e.g. DMVs, issuers) both in online and in-store settings. The platform uses advanced cryptographic techniques to enable streamlined identity check operations at scale, protects consumer privacy and reduces potential for fraud.
Related Patents/Publications:
1. S. S. Arora, S. Badrinarayanan, S. Raghuraman, K. Wagner and G. Watson, "Privacy-Preserving Identity Data Exchange based on Hybrid Encryption", US 63/149125, February 2021.
2. G. Watson, K. Wagner, S. S. Arora, S. Badrinarayanan, S. Raghuraman, B. Sullivan, D. Sloan, H. Ngo, "Integrating Identity Tokens into Interactions" US 63/115475, November 2020.
3. G. Watson, S. Badrinarayanan, S. Raghuraman, S. S. Arora and K. Wagner, "Privacy-Preserving Identity Data Exchange", US 63/104723, October 2020.
4. K. Wagner, S. S. Arora, G. Watson, S. Agarwal and M. Christodorescu, Privacy-preserving identity attribute verification using policy tokens, US 16/928,367, July 2020.
Security of Machine Learning Systems
Machine learning systems have been shown to be vulnerable to adversarial attacks in both digital and physical domains. In this research, we develop novel adversarial attacks and design defense mechanisms for detecting adversarial attacks.
Related Publications/Patents:
1. A. Ahmed, Y. Wu, S. S. Arora, Y. Wang, F. Wang, H. Yang and D. Mohasein, "Adversarial Example Detection using Latent Neighborhood Graph", In the Proceedings of the International Conference on Computer Vision (ICCV), 2021.
2. Y. Wu, S. S. Arora, Y. Wu, and H. Yang, "Beating Attackers At Their Own Games: Adversarial Example Detection using Adversarial Gradient Directions", In the Proceedings of the AAAI Conference on Artificial Intelligence (AAAI), 2021. [pdf]
3. Y. Wu, S. S. Arora, Y. Wu and H. Yang, "System, Method, and Computer Program Product for Determining Adversarial Examples", US 17/106,619, November 2020.
4. A. Abusnaina, Y. Wu, S. S. Arora, and H. Yang, "Detect Adversarial Examples By Graph Generation", US 63/088371, October 2020.
5. Y. Wu, S. S. Arora and H. Yang, "System, Method, and Computer Program Product for Determining Adversarial Examples with Adversarial Gradient Directions", US 63/075537, September 2020.
6. DL. Nguyen, S. S. Arora, Y. Wu and H. Yang, "Adversarial Light Projection Attacks on Face Recognition Systems: A Feasibility Study", In the Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Biometrics Workshop, 2020. [pdf]
7. H. Yang, S. S. Arora and Y. Wu and D.L. Nguyen, "Evaluating the Security of a Facial Recognition System Using Light Projections", US 16/752336, January 2020.
8. S. S. Arora and C. Tymoszek, "Method and System for Evaluating a Face Recognition System using a Face Mountable Device", US 16/688,871, November 2019.
9. Y. Wu, S. S. Arora and H. Yang, "Using an enrolled biometric dataset to detect adversarial examples in biometrics-based authentication system", US 16/685283, November 2019.
Biometrics System Security
One of the critical considerations for biometric systems used in real world applications, particularly in unsupervised environments, is security against spoofing attacks. In this research, we devise novel methods to spoof biometric systems, and defense mechanisms to protect biometric systems from spoofing attacks.
Related Publications/Patents:
1. Y. Zhang, M. Shirvanian, S. S. Arora, J. Huang, and G. Gu, "Practical Speech Re-Use Prevention in Voice Driven Services", In the Proceedings of the International Symposium on Research in Attacks, Intrusions and Defenses (RAID), 2021. [pdf]
2. Y. Zhang, M. Shirvanian and S. S. Arora, "Method, System, and Computer Program Product for Enabling Speaker De- Identification in Public Audio Data by Leveraging Adversarial Perturbation", US 63/148815, February 2021.
3. S. S. Arora and K. Wagner, "A Method For Iris Liveness Detection Using Passive Light Trigger", Technical Disclosure Commons, 2020. [pdf]
4. M. Leng, S. S. Arora, and K. Wagner, "Replay spoofing detection for automatic speaker verification system", US20200053118A1, August 2019. [link]
5. S. S. Arora, K. Wagner and S. Sarraf, "Defense mechanism against component wise hill climbing using synthetic face generators", US 16/356989, March 2019. [link]
6. J. F. Sheets, S. S. Arora, L. Best-Rowden and K. Wagner, "Facial anti-spoofing method using variances in image properties", PCT/US2018/041795, July 2018. [link]
Privacy-Preserving Biometric Systems
Biometric recognition is emerging as one of the key enabling technologies for validating the identity of users for both in-store and online payments. Existing payment solutions rely on either personal device or server-based biometric recognition methods. While these methods work well in practice, they typically trade off either privacy, security or convenience for the user. We design privacy-preserving biometric systems that use advanced cryptographic techniques such as homomorphic encryption and secure multi-party computation to enable secure and seamless payment experiences for users.
Related Publications/Patents:
1. C. Tymoszek, S. S. Arora, K. Wagner and A. K. Jain, "DashCam Pay: A System for In-vehicle Payments using Face and Voice", International Joint Conference on Biometrics (IJCB), 2020. [pdf]
2. E. Starr, D. F. Olson, A. Dawson, A. Jimenez, J. He, A. Sisodiya, J. A. T. P. Palacios, B. P. Parikh, S. V. Mahajan, S. Vudduraju, S. S. Tadepalli, L. Best-Rowden, K. Wagner, S. S. Arora and S. Lohtia, "Authentication Based on Biometric Identification Parameter of an Individual for Payment Transaction", PCT/US2019/013151, January 2019. [link]
3. K. Wagner, S. S. Arora and L. Best-Rowden, "Privacy protecting de-duplication", PCT/US2018/052393, September 2018. [link]
4. J. F. Sheets, K. Wagner, S. S. Arora, L. Best-Rowden and C. Jiang, "Server-assisted privacy protecting biometric comparison", PCT/US2018/043656, July 2018. [link]
5. S. S. Arora, K. Wagner, and J. F. Sheets, "Efficient hands free interaction using biometrics", PCT/US2018/044139, July 2018. [link]
6. J. Blackhurst, K. Wagner, J. Sheets, C. Jiang and S. S. Arora, "Use of biometrics and privacy preserving methods to authenticate account holders online", PCT/US2018/043872, July 2018. [link]
7. S. S. Arora, L. Best-Rowden and K. Wagner, "Distributed biometric comparison framework", PCT/US2018/023365, March 2018. [link]
Machine Learning based Authentication Mechanisms
In this research, we develop novel machine learning based mechanisms for identity document validation and biometric authentication.
Related Patents:
1. S. S. Arora, K. Wagner, J. Sheets, and L. Best-Rowden, "A dynamic learning system for intelligent authentication", PCT/US2019/014331, January 2019. [link]
2. S. S. Arora and K. Wagner, "System, method and computer program product for authenticating identification documents", PCT/US2017/054898, October 2017. [link]
Performance Measurement of Fingerprint Recognition Systems
Fingerprint based recognition systems are being increasingly used in government, forensics as well as consumer applications. To ensure reasonable performance in practice, it is very important to thoroughly test these systems before actual deployment. We develop 2D and 3D targets, as well as benchmarking methods to evaluate performance of fingerprint readers, feature extractors and matchers in the operational setting.
Related Publications:
1. J. Engelsma, S. S. Arora, A. K. Jain and N. Paulter, "Universal 3D Wearable Fingerprint Targets: Advancing Fingerprint Reader Evaluations", IEEE Transactions on Information Forensics and Security, 2018. [pdf]
2. T. Chugh, S. S. Arora, A. K. Jain, and N. G. Paulter Jr. "Benchmarking Fingerprint Minutiae Extractors" in the Proceedings of the 16th International Conf. of the Biometrics Special Interest Group (BIOSIG), 2017.[pdf]
3. S. S. Arora, A. K. Jain and N. G. Paulter Jr., "Gold Fingers: 3D Targets for Evaluating Capacitive Readers", IEEE Transactions on Information Forensics and Security, 2017.[pdf]
4. S. S. Arora, A. K. Jain and N. G. Paulter Jr., "3D Whole Hand Targets: Evaluating Slap and Contactless Fingerprint Readers", In the Proceedings of the International Conference of the Biometrics Special Interest Group (BIOSIG), 2016. [pdf]
5. S. S. Arora, K. Cao, A. K. Jain and N. G. Paulter Jr., "Design and Fabrication of 3D Fingerprint Targets", IEEE Transactions on Information Forensics and Security, 2016. [pdf]
6. S. S. Arora, K. Cao, A. K. Jain and N.G. Paulter Jr., "3D Fingerprint Phantoms", In the Proceedings of the 22nd International Conference on Pattern Recognition (ICPR), 2014. [pdf]
Biometric Recognition of Infants and Toddlers using Fingerprints
There is a growing demand for biometrics-based recognition of children for a number of applications, particularly in developing countries where children do not have any form of identification. These applications include tracking child vaccination schedules, identifying missing children, preventing fraud in food subsidies, and preventing newborn baby swaps in hospitals. This research focuses on the development of fingerprint-based recognition system for infants and toddlers (age range: 0-4 years).
Related Publications:
1. A. K. Jain, S. S. Arora, K. Cao, L. Best-Rowden and A. Bhatnagar, "Fingerprint Recognition of Young Children", IEEE Transactions on Information Forensics and Security, 2017. [pdf]
2. A. K. Jain, S. S. Arora, L. Best-Rowden, K. Cao, P. S. Sudhish, A. Bhatnagar and Y. Koda, "Giving Infants an Identity: Fingerprint Sensing and Recognition", In the Proceedings of the International Conference on Information and Communication Technologies and Development (ICTD), 2016. [pdf]
3. A. K. Jain, S. S. Arora, L. Best-Rowden, K. Cao, P. S. Sudhish and A. Bhatnagar, "Biometrics for Child Vaccination and Welfare: Persistence of Fingerprint Recognition for Infants and Toddlers", MSU Technical Report, MSU-CSE-15-7, 2015. [pdf]
4. A. K. Jain, K. Cao and S. S. Arora, "Recognizing Infants and Toddlers using Fingerprints: Increasing the Vaccination Coverage", In the Proceedings of the International Joint Conference on Biometrics (IJCB), 2014. [pdf]
Latent Fingerprint Matching
Latent prints are partial fingerprints that are left behind on surfaces of objects when they are touched or handled. They are commonly found at crime scenes and serve as an important source of forensic evidence in a court of law. Automatic matching of latent prints to rolled/plain (exemplar) fingerprints with high accuracy is quite vital for forensic applications. However, latent prints are typically of poor quality with complex background noise which makes latent print matching a significantly challenging problem. In this research, we devise methods to improve both automated and expert assisted latent matching systems.
Related Publications:
1. S. S. Arora, K. Cao, A. K. Jain and G. Michaud, "Crowd Powered Latent Fingerprint Identification: Fusing AFIS with Examiner Markups", In the Proceedings of the 8th International Conference on Biometrics (ICB), 2015. [pdf]
2. S. S. Arora, E. Liu, K. Cao and A. K. Jain, "Latent Fingerprint Matching: Performance Gain via Feedback from Exemplar Prints", IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014. [pdf]
3. E. Liu, S. S. Arora, K. Cao and A. K. Jain, A Feedback Paradigm for Latent Fingerprint Matching, In the Proceedings of the IAPR/IEEE 6th International Conference on Biometrics (ICB), 2013. [pdf]
Emerging Challenges for Iris Recognition Systems
Iris recognition technology is being used in large scale national ID systems (e.g. Aadhaar) for subject de-duplication and verification, as well as in consumer mobile phones for user authentication. Although iris recognition usually performs well in ideal to semi-controlled operating conditions, recognition performance can be adversely impacted in non-ideal or unconstrained conditions, for example, outdoor environments. In this research, we establish and solve some emerging challenges for iris recognition systems, including iris recognition under alcohol influence, iris camera interoperability and iris recognition pre and post cataract surgery.
Related Publications:
1. S. S. Arora, "Emerging Challenges in Iris Recognition", B.Tech. Thesis (IIIT Delhi), 2012. [pdf]
2. S. S. Arora, M. Vatsa, R. Singh and A. K. Jain, On Iris Camera Interoperability, In the Proceedings of the IEEE Fifth International Conference on Biometrics: Theory, Applications and Systems (BTAS), 2012. (Best Poster Award) [pdf]
3. S. S. Arora, M. Vatsa, R. Singh and A. K. Jain, Iris Recognition under Alcohol Influence: A Preliminary Study, In the Proceedings of the IAPR/IEEE 5th International Conference on Biometrics (ICB), 2012. [pdf]