EAGER: SaTC: Sweaty Digits: Bridging Chemistry and AI-Empowered Imaging for Secure and Trustworthy Human Identity Verification

与目前的生物识别技术相比，汗液可以更好地代表人类身份，具有更明显的特征，克服了现有系统的局限性，如人口统计学差异（例如，女性的准确性较低）和易受欺骗攻击。这项研究旨在通过更丰富的信号来定义人的身份，不仅是空间特征，还有相关的化学成分，由高光谱成像仪在不使用试剂的情况下捕获。本提案旨在通过高光谱成像（HSI）分析汗液来创建一种新的人类身份表征，从而使进一步的研究能够探索汗液作为高效、准确和安全的生物识别人类身份验证的解决方案。通过结合生物特征的化学特性，汗液可以提供一个细致的身份视角。通过关注汗液作为生物识别模式的HSI视角，该项目建立了更深层次的身份档案，使真人和数字表现之间的联系更加紧密，随后，系统处理它时更不容易出错，更能抵御欺骗。该项目更广泛的意义和重要性在于将化学汗液分析的进展与成像联系起来，为应用于汗液的HSI学习技术的推理奠定基础。该项目的新颖之处包括使用HSI创建汗液代谢物的档案，从而创建基于汗液的数字人类身份。为了实现这一目标，本项目着重于确认代谢物可以从人类指尖分泌的汗液中提取，确认其可重复性，并为每个感兴趣的代谢物创建HSI参考。由于传感方法的多样性，传统的汗液代谢物光谱参考不能作为HSI参考。该研究探讨了一些重要的方面，如如何获取合适的汗液样本，以及沉积的样本和捕获过程是否可重复-应用HSI。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Compared to current biometric technologies, sweat can better represent human identity with more discerning characteristics, overcoming limitations of existing systems such as demographic differentials (e.g., lower accuracy in women) and vulnerability to spoof attacks. This research aims to define human identity through richer signals, not only spatial features but also associated chemical content, captured by a hyperspectral imager without the use of reagents. This proposal aims to create a new representation of human identity based on the analysis of sweat through hyperspectral imaging (HSI), which enables further research to explore sweat as a solution for efficient, accurate, and secure biometric human identity verification. Sweat can provide a meticulous perspective on identity by incorporating chemical properties of a biometric trait. By focusing on an HSI perspective of sweat as biometric modality, this project builds a deeper profile of the identity makes the link between the genuine person and the digital representation stronger and, subsequently, the system processing it less prone to errors and more resilient to spoofing. The project's broader significance and importance is to bridge advances in chemical sweat analysis to imaging that builds foundations for reasoning on HSI learning techniques applied to sweat. The project’s novelties include creating profiles of sweat metabolites using HSI, thereby creating a digital human identity based on sweat. To accomplish this objective, this project focuses on confirming that metabolites can be extracted from sweat excreted from human fingertips, confirming their reproducibility, and creating an HSI reference for each metabolite of interest. Due to diversity in the sensing approach, spectral references obtained through traditional spectroscopy for sweat metabolites cannot be used as HSI reference. The research investigates important aspects such as how to acquire appropriate sweat samples and whether the deposited sample and the capture process are repeatable - with the application of HSI.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.