Meet Sargur Srihari
Have you ever wondered how scientists and crime scene investigators use handwriting to solve crimes? We’ve interviewed Professor Sargur Srihari about how handwriting can be used as evidence. He created an automated system that uses computers to identify patterns in handwriting and other forensic evidence. The system is used all over the world and has helped solve crimes and has been used to convict criminals!
Professor Srihari’s research led to the first large-scale handwritten address interpretation systems in the world. Postal services such as the U.S. Postal Service, Australia Post, and UK Royal Mail use his system to fight crimes! His work has also encouraged the acceptance of handwriting testimony in court, along with new software that is used all over the world by forensics specialists working to solve crimes.
Sargur Srihari is a SUNY Distinguished Professor in the Department of Computer Science and Engineering at the University at Buffalo, The State University of New York. With support from the United States Postal Service for over 20 years, he founded CEDAR, the Center of Excellence for Document Analysis and Recognition, in 1991.
Srihari: Computational forensics is the development of mathematical and software techniques to assist the forensic scientist. It’s used today for fingerprint matching in a system known as IAFIS (integrated automatic fingerprint identification system) used by the FBI and other law enforcement agencies. It includes not only fingerprints, but also criminal histories; mug shots; scar and tattoo photos; physical characteristics like height, weight, and hair and eye color; and aliases. It is the largest biometric database in the world!
Q: How does automation and computing improve traditional forensics work, such as fingerprinting, handwriting analysis, and DNA sampling?
Srihari: It allows quantitative measurement that would be otherwise difficult or infeasible. For instance it is now possible to quantitatively determine the strength of forensic evidence, such as a given configuration of small parts of a fingerprint, by stating the probability of finding it in a population of a given size.
Q: Are you surprised to find that your work is focusing on forensics?
Srihari: It was surprising that there was lack of scientific rigor in many forensic disciplines despite the techniques used being over a hundred years old. Methods used were quite subjective and susceptible to human bias. Software to solve many common-place problems has become available only in the last few decades. Techniques from artificial intelligence, machine learning and pattern recognition can play a role in performing many tasks of the forensic examiner.
Srihari: A degree in engineering is useful to develop systems for many different applications. Forensics is one of them.
Q: Is everyone using computational forensics around the world?
Srihari: Computational forensics is not yet mainstream. But lots of research goes on in academic settings, such as at my own lab at the State University of New York at Buffalo, and eventually the courts may allow these techniques to be applied in criminal trials.
Q: How long have you been a member of IEEE? What prompted you to join?
Srihari: I joined as a graduate student 40 years ago. I wanted to keep informed of new developments through IEEE publications. It was well before the internet and hard-copy magazines were the only way.
Q: What is the most rewarding thing about the work you do?
Srihari: It is learning about new things or ideas and seeing how they can be further developed.
Q: Can you share a story about how a criminal has been prosecuted as a result of the work you do?
Srihari: I was asked to testify in a federal court case about the validity of forensic handwriting comparison. It involved a government official who had claimed that her handwriting and signatures on some documents had been forged. I testified that experiments with handwriting comparison software showed that handwriting in course of business writing is sufficiently individualistic so as to be allowed in court. The defendant pleaded guilty.
Q: What advice would you give a student who was interested in working in forensics as related to engineering and computing?
Srihari: Prepare yourself with knowledge of probability and statistics which are important to all forensic specialties. While the job market in forensics is not large, that preparation can serve you very well wherever you end-up. To work with impression evidence such as fingerprints and footwear, knowledge of computer image processing will be useful.