This AI tool uses machine learning to detect whether people are social distancing properly – Mashable SE Asia

Perhaps the most important step we can all take to mitigate the spread of the coronavirus, also known as COVID-19, is to actively practice social distancing.

Why? Because the further away you are from another person, the less likely you'll contract or transmit COVID-19.

But when we go about our daily routines, especially when out on a grocery run or heading to the hospital, social distancing can be a challenging task to uphold.

And some of us just have God awful spatial awareness in general.

But how do we monitor and enforce social distancing when looking at a mass population? We resort to the wonders of artificial intelligence (AI), of course.

In a recent blog post, the company demonstrated a nifty social distancing detector that shows a feed of people walking along a street in the Oxford Town Center of the United Kingdom.

The tool encompasses every individual in the feed with a rectangle. When they're properly observing social distancing, that rectangle is green. But when they get too close to another person (less than 6 feet away), the rectangle turns red, accompanied by a line 'linking' the two people that are too close to one another.

On the right-hand side of the tool there's a 'Bird's-Eye View' that allows for monitoring on a bigger scale. Every person is represented by a dot. Working the same way as the rectangles, the dots are green when social distancing is properly adhered to. They turn red when people get too close.

More specifically, work settings like factory floors where physical space is abundant, thus making manual tracking extremely difficult.

According to Landing AI CEO and Founder Andrew Ng, the technology was developed in response to requests by their clients, which includes Foxconn, the main manufacturer of Apple's prized iPhones.

The company also says that this technology can be integrated into existing surveillance cameras. However, it's still exploring ways in which to alert people when they get too close to each other. One possible method is the use of an audible alarm that rings when individuals breach the minimum distance required with other people.

According to Reuters, Amazon already uses a similar machine-learning tool to monitor its employees in their warehouses. In the name of COVID-19 mitigation, companies around the world are grabbing whatever machine-learning AI tools they can get in order to surveil their employees. A lot of these tools tend to be cheap, off-the-shelf iterations that allow employers to watch their employees and listen to phone calls as well.

Landing AI insists that their tool is only for use in work settings, even including a little disclaimer that reads "The rise of computer vision has opened up important questions about privacy and individual rights; our current system does not recognize individuals, and we urge anyone using such a system to do so with transparency and only with informed consent."

Whether companies that make use of this tool adhere to that, we'll never really know.

But we definitely don't want Big Brother to be watching our every move.

Cover image sourced from New Straits Times / AFP.

Read the original here:
This AI tool uses machine learning to detect whether people are social distancing properly - Mashable SE Asia