Artificial intelligence taught to recognize human gait

Not everyone knows that every person has an individual style of walking. Everyone's walk is unique, like a snowflake, an eye retina or fingerprints. It is difficult for a person to see the differences, but artificial intelligence is able to learn this.

According to a new study published in IEEE Transactions on Pattern Analysis and Machine Intelligence, the new system, called StoofBD, is capable of practically unmistakably recognizing people on the gait. It is 380 times more accurate than the previously developed system and works autonomously.

StoofBD was developed at the University of Manchester. Its creator claims that during walking we use 24 different movements, from which a unique gait is formed. To create the recognition system, a database of 20 thousand steps of 120 people was created. Each step was measured and captured in high resolution. In addition, each step was measured using special carpets. Information from these carpets is processed by artificial intelligence, finding matches. This is what makes the system so accurate. 

It is assumed that StoofBD can be used for security purposes at airports and at home.

hi-news.ru