Facial recognition has been in the news a few times lately, and here it is again. As reported by us a short while ago, Chinese and UK police have used special facial recognition technology that can pick someone out of a crowd. China has been building on their surveillance network further; now with 170 million CCTV cameras in operation and plans for 400 million more to be installed over the next 3 years. But it’s not all as bad as it sounds…
Facial Recognition Sunglasses
[clickToTweet tweet=”The company behind the #FacialRecognition #technology, LLVision claims their technology takes only 100 milliseconds to scan a face and recognise a match from a database of 10,000 suspects. #AFR #AI #Privacy #TechOfTheWeek” quote=”The company behind the facial recognition technology, LLVision claims their technology takes only 100 milliseconds to scan a face and recognise a match from a database of 10,000 suspects.”]
In January/February earlier this year Chinese police showcased their facial recognition sunglasses as a trial in Zhengzhou. The aim was to catch a number of criminals and wanted persons during the New Year “Lunar” celebrations. During this time, around 385 million people were expected to travel around the country to see family and friends and welcome in the new year. As you can imagine, with this many people swarming the country, technology sounds almost a necessity to pick a face out of the crowd.
The company behind the facial recognition technology, LLVision claims their technology takes only 100 milliseconds to scan a face and recognise a match from a database of 10,000 suspects. LLVision’s CEO, Wu Fei, said, “instant and accurate feedback” is given to police, benefiting from the lack of lag traditional CCTV facial recognition systems suffer. With a built-in artificial intelligence frontend, suspects are also discovered in the very instant they come into range. This allows the officer in charge to make the next move within seconds. Unlike the CCTV systems in which the suspect is unlikely to be in that spot once discovered. The glasses cost $636 and that’s without the facial recognition technology.
So far at least 33 people have been arrested, 7 for major crimes and 26 for travelling using fake documentation. There are immediate privacy concerns in a nation that is already one of the most watched nations on the planet, however, it may take some time before the true benefits (if any) are realised. It’s certainly a topic of hot debate.
A Man Caught in a Crowd
One of the latest stories to come out about China’s use of FR technology is the news of a man who was identified whilst at a pop concert amongst a crowd of 60,000 people. Known as Mr Ao, Ao was apprehended during the concert after he was detected by facial recognition systems upon entering the stadium. Upon his capture, he made surprised remarks.
The Use of Automated Facial Recognition (AFR) in the UK
China isn’t the only nation to be using facial recognition technology. The UK has also been trialling the use of various FR methods. In December last year, Welsh police used an AFR kit during the football finals to monitor activity and check for suspects.
It is used in two ways:
- The first method uses a database of 500,000 people from their time in custody. The cameras then detect a face in the crowd, cross-reference that image against the ones in the database, and if it’s a match the officer is alerted is sent to apprehend the wrong-doer.
- The second method uses a more ‘live’ approach. For example, someone gets violent whilst on a night out and it’s captured on CCTV. This footage is then checked against the CCTV film of the current day, and a match is made this way instead.
Of course, like with any technology in its infant stages AFR has its issues. London police, whilst using the facial recognition technology at Notting Hill Carnival, reported 35 false matches and one wrongful arrest. It’s clear that the technology needs some refinement before it should be heavily relied on. Especially when there are already concerns over privacy infringements; arresting an innocent party is likely to spark controversy.
How Does Facial Recognition Technology Work?
So what does the tech behind the technology look like? FR uses an algorithm to determine certain characteristics of the human face. Characteristics include the nose, jaw-line, mouth, ear-size and distance between the eyes. The application would see a subject move in front of, or pass by, a facial recognition-enabled device in which a sample of their physical appearance is extracted. A ‘template’ is then created from the biometrics and compared to another source. This is perhaps where the worry around privacy comes in. A photo of us would need to be kept in a database and that would make many feel uneasy. With the upcoming GDPR regulations we could see our data kept a lot more securely, thus easing some worries, however, the risk of breach and misuse will always be ever-present. The final step is the system’s decision as to whether the images greatly compare and a match is found.
Let’s take a look at situations in which FR technology can be used in a way that benefits us in our day to day lives.
Non-Legal uses of FR Technology
Let’s end things on a slightly more positive, optimistic note and look at some potential (and in some cases, already implemented) uses of FR technology:
- In restaurants – When paying for a meal
- At a cash point – To save time withdrawing money
- Government offices – To only allow those with appropriate security clearances in
- Airports – When checking-in or going through immigration
Next in the series, Wednesday 9th May 2018…
Tech of the Week #9: Medical drones supplying life-saving treatments
Drones are used for many reasons, both business and personal. Amazon can deliver your new pair of shoes and you can get a great aerial shot of your favourite chill out spot. What if I told you there are medical drones that deliver life-saving treatments such as blood?
If you want to talk tech or facial recognition, and for more great content, pick my face out of the Twitter crowd @JoshuaOsborn16.
I work for Compare the Cloud as programme manager. I enjoy cybersecurity, fintech and, on the less boring side of things, photography, trains (I said less boring, right?) and, like everyone else, music.