Sandip P

Research Assistant @Clarkson University Graduate School

Potsdam, NY, US

Presently, the project lead in, Developing noncontact-fingerprint-based presentation attack detection (PAD) models Generating high-quality, synthetic noncontact-fingerprint-based fingertip images Developing noncontact-fingerprint-based image segmentor to aid data processing Developing statistical methods for biometric system performance bias detection across demographic user groups Skills: CNN architecture-based classification, StyleGAN, Keras, TensorFlow, Image Processing• Investigated the impact of aging on voice recognition in children. The results have direct benchmark applications in industrial and government forums involved with voice biometric applications in children.Skills: Statistical data analysis and non-linear modeling• Participated in the preparation, presentation, and implementation of four successful research grants - approx.$200k in total• Project Lead in two NSF-CITeR funded projects- LivDet: Liveness detection competition – an international competition series evaluating the state-of-the-art PAD solutions: Face 2021 and Non-Contact Finger- 2022 (upcoming). Involved in hosting LivDet-Face 2021 competition, Co-organized LivDet-Iris 2020 competition, developed presentation attack datasets, analyzed algorithms, and published reports on the performance of face and iris PAD algorithmsResponsibilities: Organizing, designing collection protocol, data collection, cleaning, and organizingSkills: Management, data handling, and interpersonal skills.• Experienced in the development and analysis of a novel cardiac radar system for measuring heartbeat from a distance, along with system capacity estimation for continuous biometric verification