Researchers have developed a behavior-based biometric verification system that provides a non-intrusive authentication method by analyzing the way individuals walk and their gait, which can be used for airport security to replace fingerprint recognition and iris scanning...
The Department of Electrical and Electronic Engineering at the University of Manchester in the United Kingdom collaborated with researchers at the University of Madrid in Spain to develop a behavioral biometric verification system that measures human gait or walking patterns, as long as individuals walk The pressure pad on the floor can analyze the 3D form and time-based data of the footsteps to successfully identify the individual.
Researchers claim that the system analyzes the way individuals walk and their gait, providing a non-intrusive authentication method that can be used as a biometric technology for airport security to replace fingerprint recognition and iris scanning.
The results of the study were published in the Journal of Machine Learning Research at the beginning of this year, IEEE TransacTIons on Pattern Analysis and Machine Intelligence (TPAMI). The researchers pointed out in the article that, on average, this newly developed artificial intelligence (AI) system correctly identifies individuals almost 100% of the time, with an error rate of only 0.7%.
Currently, biometric-based biometrics such as fingerprinting, face recognition, and retinal scanning are often used for security purposes. However, behavioral biometrics such as gait recognition can also capture unique identifiers that are presented by individual natural behaviors and patterns of action. The research team used a number of so-called "impersons" and a small number of users to test their data for three real-world security scenarios - airport security checkpoints, workplaces, and home environments.
Omar CosTIlla-Reyes, professor of electronics and electrical engineering at the University of Manchester, who led the study, explained: “Everyone is affected by about 24 different factors and actions while walking, resulting in a unique and unique gait pattern for everyone. As with fingerprints or iris scans, monitoring these actions can be used to identify and verify an individual's identity."
In order to build such an AI system, the computer must be allowed to learn these modes of action. The research team collected the largest gait database to date, including nearly 20,000 footstep signals from 127 individuals. To compile these samples and data sets, the research team also used floor-standing sensors and high-resolution cameras.
This data set is called SfootBD. CosTIlla-Reyes is using this SfootBD dataset to develop the advanced computational models required for automated footprint biometric systems published in TPAMI.
Gait modes include vision, pressure, and accelerometers.
CosTIlla-Reyes added: "It is very challenging to study non-invasive gait recognition by monitoring the strength of the individual's footsteps on the floor. This is because it is extremely difficult to distinguish between subtle differences between people. That's why we have to develop a new AI system that addresses this challenge from a new perspective."
One of the main advantages of using footstep recognition is that the entire monitoring process is non-invasive to the individual and is adaptable to the surrounding noisy environment, compared to the shooting or scanning method used at the airport security checkpoint. At the same time, when stepping on the pressure pad for monitoring, people don't even need to take off their shoes, because the system relies not on the shape of the footprint, but on the gait of walking.
Other applications of the technology include a smart step that identifies neurodegeneration, which will have a positive impact on the healthcare sector. At the same time, this is another area where Costilla-Reyes intends to advance its research through footstep recognition.
He added: "This research is also in progress. It is expected to address the health care issues that identify cognitive decline and mental illness episodes through raw footsteps from sensors deployed in vast floors of smart homes. Human movements may become another novel biomarker for cognitive decline and unexplored through new AI systems."
Engine Parts oil tube,Engine Parts water pipe,Engine Parts exhaust pipe,Engine Parts Rubber Pipe
Chongqing LDJM Engine Parts Center , http://www.ckcummins.com