As organizers of the largest innovation competition in the security world, one of the Global Security Challenge’s tasks is to map emerging innovations, such as 2nd generation biometrics. Below is a quick overview of where we stand with 2-gen biometrics such as typing behavior, gait analysis and odor sensing.
TYPING
I would describe typing biometrics as the 21st century version of the good ‘ole signature analysis. Today’s typing algorithms have been developed by converting a person’s typing attributes such as speed, rhythm, agility, corrective behavior and use of shift keys into a unique identifier. These systems are pretty advanced with heir inventors claiming their software to be keyboard-independent and continuously adapting to slightly changing typing habits. There are a couple weaknesses that remain including for example the time difference it takes to type a sentence instead of a short 6-letter password onto your PC’s log-in screen. In addition, its accuracy levels are still disputed by some government experts with whom we discussed this technology. Nevertheless, this technology is on a fast track to maturity and could have tremendous impact on commercial and intelligence sectors.
GAIT ANALYSIS
The University of Southampton has positioned itself as one of the front runners on gait-biometrics. Their prototype identified people by the way they walk with a success rate of over 90%, under perfect lab conditions. According to Computing.co.uk, the university recently built a tunnel-like system that consists of eight digital cameras, which register persons’ movements and generates 3D-models of their walking behavior. See below for a graphic representation of such a modeling, based on Professor Nixon’s research
Having reached high accuracy rates in their lab, the university is currently planning to take its system from the lab into the real world to test it under real-life conditions. Such real-life tests will be helpful to stress-test popular challenges against gait-biometrics, such as whether our walking style is affected by illness, injury or even just stress. For instance, will the system accurately recognize me while running to catch my plane as well as casually strolling through the airport lounge?ODOR SENSING
The idea of identifying people by their smell is nothing new as we humans apply it daily (subconsciously) and even use it to find criminals taking advantage of dogs’ advanced noses. The attempt to create an electronic nose, dubbed in the industry as ENose is by creating a chemical sensor that detects the small quantities of molecules that evaporate from our body. The limitations of the current systems is that the human nose has over 10.000 sensors to detect these molecules that are processed by about 10 million sensory neurons in our brain, while ENoses in contrast operate with only 10-20 sensors and proportionally smaller number of artificial neurons. (perhaps we ask Steven Jobs to create an iNose?)
The current state of odor-sensing technology is nicely summarized by Daniel Lee, a bioengineering scientist at the University of Pennsylvania, who said in 2003: "We have cameras that can see outside the spectrum of the human eye and microphones that can detect a vibration a mile away, but in terms of chemical sensing, we are far away from what biology can do." Below is a picture of a prototype of electronic nose, by the University of Pennsylvania.
2nd Generation Biometrics – An Endangered Species
It has taken some of these technologies over a decade to get from research to a working technology, so it is crucial for disruptive technologies, such as 2nd generation biometrics, to have enough time to improve accuracy and reliability. Thus, funding is crucial now for them to overcome this first valley of dearth on the road of commercialization. An alternative to external funding is to first target commercial clients with not-so sensitive data to protect, which in addition to bootstrapping the company would act as a real-life testing bed for their beta technologies.
No comments:
Post a Comment