A London start-up is developing CCTV cameras that can 'read' emotions of people in crowds

p29 EDITION 16.12.19 PIXELATED Crowd Recognition CCTV CCTV cameras can now be customised to "see" if large crowds in station stations or at football matches are becoming angry. A London start-up has developed artificial intelligence-powered computer chips and software that scans faces and bodies to predict human behaviour after analysing millions of source images of how groups of people act in public spaces. The Shoreditch-based firm, Sensing Feeling, said the huge numbers of human actions scanned during research means its technology could help detect if someone on the platform is calculated to be at risk of harming themselves. Their "risk prediction" software, which can be installed into existing CCTV networks, uses computer vision to recognise human behavioural patterns to assign crowds a real-time "emotional index". MFL

A tech firm says it has developed software that enables CCTV cameras using artificial intelligence to “read” the emotions of people in crowds.

Soho-based start-up Sensing Feeling says the system has the potential to predict people’s behaviour by scanning faces and body language to assess their mood. The footage is checked in real-time against pre-programmed images for signs of anger, contempt, fear, disgust, happiness, sadness and surprise.

Developers say the “risk prediction” technology has wide-ranging applications in industry, retail, education, arts and the workplace.

It could warn when workers become tired, check the effectiveness of in-store displays or tell a restaurant boss whether or not guests are enjoying their meals.Rail staff would be alerted when rush-hour crowds become angry or frustrated.

It is already in use at a private university outside the UK which licensed the system to replace student satisfaction feedback forms after lectures.

Jag Minhas, Sensing Feeling founder, told the Standard the firm was alive to concerns over privacy and ethics, and said safeguards were in place to ensure it complies with data laws.

A Crowd Risk Index can be calculated by scanning faces and body language

According to the invention’s European patent, the software maps facial characteristics, upper and lower body poses, hand gestures and interactions with others. It uses computer vision to recognise behavioural patterns and assign crowds a real-time “emotional index”, shown up as on-screen percentage predictions or as heat map-type colours

Mr Minhas said: “Our technology can sense how people are feeling, just by observing their behaviour. It can assist in the more effective management of that by highlighting risk associated with crowding in public spaces, or behaviour that might cause accidents and near misses. A pattern of behaviour that might be associated with the risk of self-harm could also be something that this type of solution can detect.”

Despite the safeguards, this field of technology has raised serious concerns among campaigners who say existing laws are rapidly being outpaced by the speed of development. Silkie Carlo, director of Big Brother Watch, said: “This kind of surveillance aims not to monitor your physical movements but your mental state which is a profoundly dangerous concept.”

Sensing Feeling is supported by tech accelerator Wayra UK.