3D detectors measure social distancing to help fight Covid-19 

Alexandre Alahi (left) and Lorenzo Bertoni (right). © Alain Herzog / 2021 EPFL

Researchers at the Swiss Federal Institute of Technology Lausanne (EPFL) have repurposed an algorithm they initially developed for self-driving cars to help people observe social distancing rules.

“When Switzerland went into lockdown last year, we were working on an algorithm for self-driving cars,” says Lorenzo Bertoni, a PhD student at EPFL’s Visual Intelligence for Transportation Laboratory. “But we quickly saw that by adding just a few features, we could make our program a useful tool for managing the pandemic.” 

EPFL’s 3D detector, called MonoLoco, works with a camera and could even be operated on a smartphone. It detects whether individuals are maintaining  the right distance to prevent infection, without collecting any personal data, by calculating the dimensions of human silhouettes.

The researchers published their work in IEEE Transactions on Intelligent Transportation SystemsExternal link this week and will present it at the International Conference on Robotics and Automation in June.

Bertoni, the lead author, says they wanted to develop a high accuracy detector that “wouldn’t mistake a streetlight for a pedestrian.”

Detectors typically locate individuals in the 3D space by assuming they’re on the same flat surface, which requires the camera to be perfectly still and fails to catch someone going up a stair, he notes.

The EPFL algorithm also breaks new ground by identifying people’s body orientation and can determine how people are interacting, including whether they are talking. 

MonoLoco keeps the faces and silhouettes of people who are filmed completely anonymous because it measures only the distance between their joints.