Face Recognition can Now Work in Utter Darkness
This talk will present a new advancement made in enabling face recognition technology to work in darkness. The technology identifies a person from their thermal signature and matches with ordinary photos. Cross modal face or general image matching between the thermal and visible spectrum is a much desired capability for night-time surveillance and other civil/security applications. Due to a very large modality gap, thermal-to-visible face recognition is one of the most challenging matching problems.
In this talk, we will present new technology based on deep neural networks in bridging this modality gap by a significant margin. While showing its effectiveness on a much harder face recognition application, the new advancements made can be easily adapted and applied to general thermal-visible image matching, potentially opening doors to many new and exciting applications. We will discuss interesting directions that can enable this technology to work in real-time and with minimal computational resources making it very practical for various applications. We will also discuss some associated privacy and legal aspects concerning the use of this technology.
Saquib Sarfraz is currently working as a senior research scientist & lecturer with the computer vision for human interaction (CV-HCI) lab, Karlsruhe Institute of Technology (KIT), Germany. He is the team lead for the work in the direction of face recognition and also member of several related funded projects, where he is working on facial analysis & tracking for identity, age, and gender classification. He obtained his PhD in Computer Vision from Technical University Berlin, Germany in 2008. Before moving to KIT, he served as assistant professor (2009-2012) at the electrical engineering department of COMSATS Institute of Technology, Pakistan. There he founded and directed the Computer Vision Research Group (COMVis). His research interests include video surveillance, forensic image analysis, face recognition, multi modal biometrics, and general machine learning.