I-Corps: Translation Potential of Contactless Biometric System for Secure Recognition

About This Grant

This I-Corps project is based on the development of a new class of biometric technologies that securely identifies individuals without physical contact. Current biometric systems, such as fingerprint or facial recognition, often face difficulties in everyday settings due to changes in lighting, user appearance, and other environmental factors. These challenges result in errors and delays, creating security risks and inefficiencies in critical areas such as healthcare, finance, government, and infrastructure. This solution applies advanced machine learning techniques to improve the way biometric systems learn and recognize unique features, increasing accuracy, reliability, and scalability for broad use. The technology performs effectively in real-world environments, enabling fast and secure identity verification without requiring physical interaction. The solution addresses the growing problem of identity theft and unauthorized access, which impacts millions of individuals annually. By reducing these risks, the technology enhances public safety, protects sensitive data, and increases operational efficiency. The technology also lowers the need for manual identity checks, saving time and resources. The project advances national interests by fostering scientific progress in secure digital identification, supporting economic stability, and strengthening infrastructure essential to public welfare. This I-Corps project utilizes experiential learning coupled with a first-hand investigation of the industry ecosystem to assess the translation potential of the technology. This solution is based on the development of a new class of biometric algorithms that leverage advanced, generative, artificial intelligence-based data augmentation pipelines; deep biometric feature learning techniques using supervised and self-supervised learning; and sophisticated decision post-processing frameworks utilizing decision uncertainty estimation techniques. These innovations improve the accuracy, robustness, and scalability of visual contactless biometric systems, especially in uncontrolled operational environments. The approach overcomes challenges related to environmental variability, user presentation differences, and large-scale deployment by integrating adaptive learning models capable of generalizing across varying environmental conditions. This solution advances the state of the art by introducing advanced data augmentation, feature extraction, and decision post-processing pipelines, thereby enhancing reliability in dynamic, real-world applications. Users benefit from highly reliable identity verification systems with faster processing times and reduced false acceptance and rejection rates. The technology supports secure, user-friendly, and contactless authentication, ideal for sectors such as healthcare, finance, and government. It contributes to the broader field of biometric solutions that enhance universal accessibility and operational resilience. Broader adoption of this innovation may reduce identity fraud risks, limits unauthorized access, and improves the efficiency and security of authentication workflows across a wide range of applications. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.