Editor's Comment - Back to school
20 June 2011 20:27 GMT

Work hard, play hard - biometrics in good hands

For those of you that haven't been to University for a while, I can tell you that students – or at least pattern recognition/biometric/forensics students – are now expected to work hard…Very hard.

I have just returned from an amazing week on the beautiful island of Sardinia, just off the Italian coast. A perfect venue for…sitting in a lecture theatre whilst 18 of the world's top biometric and forensics academics and researchers share their knowledge in individual 90-minute sessions over the course of a week.

The brainchild of Massimo Tistarelli, from the Computer Vision Laboratory at the University of Sassari, Italy, the 8th Summer School for Advanced Studies on Biometrics and Secure Authentication, was stimulating, informative enjoyable and a must-attend if you research biometrics in any shape or form (whether at University, in a company, or within a government department).

So what did I learn? Well apart from the fact that my mathematics skills are no longer up to the job, I learned that the industry's future, from an academic standpoint at least, is in excellent hands.

I had previously thought that research is too often far removed from the necessities of business. But now I have changed my mind. While it is clear that working at the theoretical forefront of pattern recognition science will not always produce immediate results, it is the body of learning that builds upon it that will - from time to time - unearth a new concept or technique that makes a difference. Advancements in facial recognition and speaker verification are testament to that.

I also heard time and again the importance of two areas – standards (including the real-world research opportunities that emerge from their development) and testing. NIST's testing of various technologies over the years were time and again lauded at the event, and it is clear that the competitive and collaborative approach they have fostered has driven forward algorithm accuracy and their tolerance under difficult environmental conditions.

Multi-modal fusion (and I can now recite perfectly the various different models that are possible and popular) also seems to be a no-brainer thanks to its ability to address the technology's inherent weaknesses – indeed human/machine fusion was one area that particularly surprised me in how it can improve results.

Of course in a whole week of lectures there is much that I learnt - from standards terminology to the possible demise of Doddington's Zoo (more of that in a future edition). I will need some months to digest the full content of the lectures.

Today's students do know how to work hard. But, I am delighted to say, that they have also retained the ability to party hard too! But perhaps that is what you get when you hold a summer school on an Island in the Mediterranean!!!
 

Mark Lockie