From [HERE] For the last two years, British authorities have been testing AI-powered emotion recognition cameras on travelers in train stations, according to recently disclosed documents.

What can the cameras see?

In each of eight train stations around the UK, five to seven cameras or sensors scan the passing crowds. Using AI software from Amazon, the cameras use object recognition algorithms to spot criminal activity, trespassing, weapons, wet floors, overcrowding, and overstuffed trash cans. 

They can also alert authorities when passengers engage in “unusual behavior,” such as shouting, running, smoking, or skateboarding. 

A combination of existing cameras and newly installed smart cameras are being used for the surveillance project. All are equipped with Amazon’s Rekognition surveillance system.

Emotional recognition

Through a combination of facial and behavioral analyses, the Rekognition software allows officials to also analyze visitors’ emotions. Internal documents note that the cameras can detect whether a passenger is “happy, sad [or] angry.” They can also recognize passenger satisfaction, which “could be utilized to maximum advertising and retail revenue.”

Anti-social behavior

Authorities can also detect passengers who exhibit “anti-social behaviour,” which according to London’s Metropolitan Police Service means “behaviour by a person which causes, or is likely to cause, harassment, alarm or distress to persons not of the same household as the person.”

Social distancing, face masks, and demographics

Other use cases for the surveillance cameras include detecting whether passengers are keeping to social distancing rules and wearing face masks. The cameras can also scan for demographics by identifying people’s age profiles and “socio-economic classification,” which the documents note can create “revenue opportunities for better targeted advertising and product placement.”