Wearable tech that can capture the full body has been developed.
Researchers at Cornell University have developed a wearable wrist camera that can capture 3D renderings of the entire human body. Called BodyTrak, the technology uses a single, tiny RGB camera and a "customized deep neural network," the university shared in a media release on Tuesday.
Once the camera captures images of the body, integrated AI technology can create full 3D silhouettes using various positions of 14 joints in the head, torso, arms and legs. Because AI sensors can mold the images by reconstructing only pieces of what the camera captures, there's no need to capture full shots of the whole body. During their study, researchers used different camera settings and followed participants as they conducted activities such as walking, working out or sitting. Based on BodyTrak's ability to detect the human figure in multiple poses and movements, they hope the technology can be used in smartwatches of the future.
"Since smartwatches already have a camera, technology like BodyTrak could understand the user's pose and give real-time feedback," said Cornell's assistant professor of science Cheng Zhang. "That's handy, affordable and does not limit the user's moving area."
The researchers aim to have BodyTrak assist with monitoring physical activity and body functions. As far as privacy concerns, they say that the camera points directly at the user and only acquires partial images of the body. Described as a dime-sized camera that sits on the wrist, the device is one of the latest developments in Cornell's SciFiLab, a facility that has also built sensing that track hand and finger movements and facial expressions.
First published on Nov. 9, 2022 at 7:11 a.m. PT.
