Biometrics: Past, Present and Future

Slides (pdf)  Size: 2.93MB

Abstract

The origins of present-day biometric recognition can be traced to the Habitual Criminal Bill passed by the British Parliament in 1869. The bill proposed registration of criminals in order to recognize previously convicted offenders when brought up for trial. In response to this bill, the anthropomorphic-based Bertillon system was developed in 1879 and later on, a fingerprint identification system was adopted by the Scotland Yard in 1905. However, it was not until the 1960s that algorithms for automated comparisons of fingerprints, face, and voice were first published (Daugman’s iris recognition algorithm appeared in the IEEE Trans. PAMI in 1993). Automated Fingerprint Identification Systems (AFIS) were later developed for use by law enforcement agencies in 1980s. Today, law enforcement agencies around the world continue to routinely use fingerprints to reliably determine the answers to the questions, “who is this person?” and “have we seen him before?” Outside the realm of forensics, biometrics have been adopted to manage the world’s largest, successful, civil registration system (India’s Aadhaar). Aadhaar has enrolled about 1.3 billion residents using their face, fingerprints and irises for de-duplication (is this person already enrolled in Aadhaar?).

The September 11, 2001 terrorist attacks in the United States shifted the focus of biometrics from law enforcement and forensics to immigration control, access to secure buildings, corporate networks, mobile phones and ATM machines, and financial transactions, to name a few. The acceptance and accuracy of biometric deployments in these applications have caused many in the broader computer vision and machine learning communities to form a premature and incorrect opinion that biometric recognition is a solved problem. In my presentation I will attempt to look at my crystal ball and address where do we go from here? and what are the most fruitful directions for biometric researchers, especially those in academia?

Biography

Anil K. Jain is a University Distinguished Professor at Michigan State University. He has received Fulbright, Guggenheim, Humboldt and King Sun Fu awards and served as the editor-in-chief of the IEEE Trans. Pattern Analysis and Machine Intelligence. He was elected to the National Academy of Engineering, The World Academy of Science, Indian National Academy of Engineering and the Chinese Academy of Sciences. He currently serves on the Board of Trustees of Mohamed bin Zayed University of Artificial Intelligence (MBZUAI), Abu Dhabi.

Anil Jain
Michigan State University, US