GAIT – AI’s NEW FRONTIER
AI is breaking new grounds by the day. With the total quantum of AI research unknown we are likely to get regular surprises. There was news sometime back that the car steering wheel would unlock once the sensors on the driver’s seat would recognise the genuine driver / owner. This sounded like sci-fi. Scientists have now developed an AI tool to identify people by measuring their gait or walking pattern.
But despite more than three decades of research, using gait analysis as a biometric did not take off, until now. Artificial Intelligence based recent advances in the accuracy of automated gait recognition clearly suggests that this technology may on the verge of being used in the new generation of security systems. It might be used replace fingerprinting & eye-scanning in airport security. Current AI based systems can successfully verify an individual simply by analysing the footstep 3D & time based data.
This system has been developed by University of Manchester, UK & University of Madrid, Spain. The strike seems to be nearly 100 % of the times, with an error rate of just 0.7%. Presently, fingerprints, facial recognition and retinal scans are used for security purposes in different combinations depending on the requirements. The team has tested the AI Gait Recognition Technology using a large number of ”imposters”& small number of real time users in three different real-world scenarios; airport security checks, the workplace & the home environment.
”Each human has approximately 24 different factors and movements when walking, resulting in every individual person having a unique, singular walking pattern.” Monitoring these movements gives a unique gait, with the same uniqueness of a fingerprint or a retinal scan. It is also called as the non-intrusive gait recognition. Extended further, it can also be used for neuro degeneration, which could have positive uses in the health care sector.
GAIT RECOGNITION CAN BECOME AN IMPORTANT SECURITY TOOL IN INDIVIDUAL IDENTIFICATION.