The forensic disciplines which are based on pattern evidence, such as fingerprints, handwriting, shoeprints, etc, are under a great deal of criticism in the courts due to several missing elements in their scientific underpinnings. The elements include: rigorous quantification of individuality, definition and establishment of error rates, publication in standard scientific journals, reproducibility of results, contextual bias, etc. At the same time the availability of easily available computing power makes it possible to take a fresh look at most of these issues so that what has been intuitively observed over a hundred years of forensic practice can now be re-established rigorously. In this talk I will present approaches to several of these issues including discussion of recent results on computational approaches to handwriting, fingerprints, and shoeprints. A demonstration of a software system to compare signatures and extended writing will also be given.

Bio:
Sargur (Hari) Srihari is a SUNY Distinguished Professor at the State University of New York at Buffalo. He teaches and does research in pattern recognition and machine learning in the Department of Computer Science and Engineering. He is presently serving on a U.S. National Academy of Sciences committee on ³Identifying the Needs of the Forensic Community². An author of over 300 papers he has received several honors including Fellow of IEEE, Fellow of the International Association for Pattern Recognition (IAPR) and Distinguished Alumnus of the Ohio State University College of Engineering.

Srihari received a B.E. from the Indian Institute of Science, Bangalore and a PhD from the Ohio State University, Columbus.

Sriharis homepage.