Introduction
So did George Orwell actually predict the future of the Cyber Age with his 1984 book?In his book, George projected a state which observed its citizens, and where there was no hiding place. He could see a time of TVs on the wall which could talk to you, and for the citizens to be watched for everything that they did. Perhaps he could see a world where our Cloud Service providers ... Google .. Facebook ... and so on ... continually monitoring our activities?
Obviously, in 1948, when he published it, the use of technology was not quite developed as it is now, but perhaps we need to examine our current move towards the observation of crime through technologies such as face recognition. On the back of Internet records in the UK being monitored without a warrant, we now see that citizens may not be free from observation when they attend even something like a football match.
Monitoring faces
A number of police forces in the UK have been trying face recognition in public spaces, including South Wales Police, London's Met and Leicestershire. Over the past year, South Wales Police have used it in 10 events. One test event involved scanning a crowd over over 35,000 conference attendees at an Elvis conference in Wales, and where 17 faces were identified as a threat, and where 10 were true-positives and seven were false-positives. It was also used at the UEFA Champions League Final Wales in June, 2017 identified 2,470 alerts of which 2,297 were false positives and only 173 were true positives. At a boxing match - Anthony Joshua versus Kubrat Pulev - there were five true-positives and 46 false positives, and at a Wales v Australia rugby match, there were only six true positives, and 42 false positives.
It has also been used within the Download music festival, at the Notting Hill Carnival, and at a remembrance service held at the Cenotaph.
Monitoring a Scottish Football matches
Recently the SPFL (Scottish Premier Football League) announced that they would seek support from the Scottish Government on the usage of facial recognition at football grounds. The worry, of course, is where you stop, as there are crimes committed everywhere, and football matches possibly do not match the threat of a Friday evening in a big city?The cost of the system is proposed at £4 million, and is intended to enforce measures related to the Offensive Behaviour at Football and Threatening Communications (Scotland) Act 2012.
We all know the risks of airline travel, and we accept that we are being monitored, as it will protect our safety, but to move to the point that everywhere we go, here there might be a criminal incident, and for our face to be scanned, is something that needs debate. Imagine you walk out your door, and the CCTV camera recognises you and logs your location, then you do to the cinema, and it logs you, and everyone you come in contact with?
How does it work?
Face recognition software has come a long way in the last few years and most systems aim to find keys points around the face and measure them:In the following, we can see that the face processing methods are able to pinpoint these key positions in the face, along with identifying age, gender, and emotion [Try here]:
The technology has now even advanced to monitoring emotion within real-time video streams:
Apart from detecting known agitators within a crowd, one application of the technology is to detect anger within crowds [Try here]:
And it's not just law enforcement that is rushing to use emotion detection, it is also the perfect metric for advertisers who want to understand the emotions of their users when they see their content:
Conclusions
There are many questions that would have to be asked. Is it worth the risk to privacy? How long is the data kept for? Will the metadata be used for other purposes (such as for a future criminal investigation)? Will it be used to link people and objects together? What will stop the data from being used by those who have malicious purposes?So how does this differ from CCTV? Well using face recognition and object identification it is not possible to extract metadata from the video capture, and use this to create linkages and to search across many sources of data.
In academia, a key focus is to, at least, debate things and open them to review, and the usage of face recognition for law enforcement is something that really hasn't been debated in any reasonable way. Face recognition is a powerful technology, and it has advanced to the point that it can identify people from remote cameras. Like it or not, there's a whole lot of cloud service agents who are also watching you, and continually tracking you, and feeding you content that you might like.
I'm a technologist, and not a politician, but this move to observing people at sporting events really needs, at least, some debate about the risks involved. While terrorism-related observation is possibly acceptable in some circumstances, the matching of faces within a sports match is something that those being observed need to have some say into the gathering of data, especially in how this data could be used for other purposes.
The nightmare scenario is that each person could be observed, and every link they have with every person and every object could be gained. Like it or not, Google has a good deal of this information, but for it to be used in a law enforcement content needs to be handled correctly, and the systems that would be used should be open to scrutiny.
Comments
Post a Comment