innovator

Add idea


Calendar

«    June 2018    »
MonTueWedThuFriSatSun
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
 

 

Advert

 

Payment

 

Advert

 

Authorization

Стартап

DateDate: 4-06-2018, 06:47

Researchers want to use this method to replace passwords, keys and tokens. All these things can lose or steal, but the steps of a person are individual. Therefore, with their analysis and modern technologies, it is possible to create doors and locks that open when a person approaches it.
British and Spanish scientists have taught the neural network to recognize a person by his gait. This technology is based on a method of in-depth study. As a result, it will be possible to determine the identity of a person with 100% accuracy for a number of individual characteristics.
A person's walk can eventually become an excellent identifier. This characteristic is divided into spatial and temporal: the first of them studies the length of the step, the turn of the leg and its position, and the second - the supporting and motor phases of the step. Also, artificial intelligence remembers individual characteristics, which makes it impossible for an attacker to forge it.
Scientists from the University of Manchester, led by Omar Custill-Reyes, collected for the database more than 20 thousand photos of tracks of 120 people. Volunteers were asked to wear any footwear that was convenient for them, which complicated the neural network task.
To learn AI all the images were downloaded, and the system analyzed them. As a result, based on their method, experts even learned to predict the behavior and mood of the dog by its gait.
The new system was tested on three datasets with different sizes. Recognition system checked in the conditions of the airport, home and workplace. As a result, the efficiency, depending on the dataset, was from 92.9 to 99.3%.
According to researchers, collecting such tracking devices is much easier than collecting a lot of fingerprints of workers, for example, large corporations. However, it remains unclear how neural networks work in abnormalities, such as fractures and sprains.
Source: VistaNews