Artificial Intelligence

Deep learning facial recognition fights human trafficking

25th September 2018
Alex Lynn

The subject of human trafficking is a gloomy topic to discuss as it is a dire problem that the world continues to face. This rampant global issue displaces children and teenagers. Trafficking is a horrifying matter that frequently transcends gender, race, socio-economic environments and social status. It can happen at any time and anywhere, devastating families and communities.

It’s important feel safe where we shop and frequent often, particularly as parents with children whereby we know their safety and security is attended to. We want our malls, parks and public facilities to be equipped with the right technology, that should these horrific kidnapping incidents occur, they’re tackled timeously, allowing for peace of mind.

Based on AxxonSoft’s Safe City projects, a concept that was developed to help countries target various types of crime, vandalism and terrorism, the company’s facial recognition and forensic search technology is able to assist the war against human and child trafficking. This software incorporates Artificial Intelligence (AI) which utilises powerful machine learning-based techniques to emulate how the human brain operates, and can work to help authorities identify unsolicited behaviour and potential human threats in public spaces such as train stations, malls, parks, stadiums and airports to name a few.

In a real life environment, control rooms can easily upload a picture, snapshot or photo-fit to quickly find suspects or victims, and related video episodes that may be of interest to police in trafficking hotspots and highly dense public areas, often the prime location for kidnapping. Axxon powered control rooms can pinpoint where and when the person of interest (victim or suspect) appeared, and what he or she did in those places. It even enables the capability to tag and track a suspect through a complete vicinity between different camera systems such as with Axxon’s Homeland Security where authorities locate places most frequented by a suspect and/or victim and discover where they were last seen.

This advanced AI technology can also automatically pick out faces from live video feeds, and trigger an alarm system when it determines a given degree of similarity, high or low. This type of surveillance and security system is considered to be a proactive approach to the trafficking problem, one which is often very reactive in nature. Deep Learning can be crucial when trafficking syndicates quickly move their victims from the scene of the crime to a safer, ‘surveillance-free’ hideout.

Furthermore, authorities are often alerted to human and child trafficking incidents after the illicit activity has taken place, and the tracking down of valuable information in an abundance of video data can be a daunting task. With Axxon’s forensic video search, authorities can instantly check data based on the criteria they are looking for, such as finding any movement in a specific area of the frame, or finding all the red cars that crossed a certain line moving in a selected direction or even as specific as locating two suspects who are heading in a known direction, one of them wearing a green jacket. As a result, what would have taken someone hours to shift through, only takes a few seconds to locate, allowing for efficiency and effective reaction to the situation.

Global Marketing Director for AxxonSoft, Colleen Glaeser, said: “If we are to try and combat or even reduce human and child trafficking around the world, drastic measures need to be implemented rather urgently. We all want a safer city, a safer country and safer world for our children to grow up in. We don’t want to be ridden with fear that the threat of trafficking may affect us directly or people we know and love. AxxonSoft’s Deep Learning facial and license plate technology is truly revolutionary in its ability to effectively and proficiently track trafficking syndicates and their victims.”

Product Spotlight

Upcoming Events

View all events
Latest global electronics news
© Copyright 2024 Electronic Specifier