Smart glove uses stretch-sensing tech for highly accurate hand gesture capture
Getting real, physical movements accurately depicted in the digital world remains a challenge for programmers and engineers, but a newly developed glove promises to advance the tech significantly. It's able to capture hand movements with much more detail and nuance than most existing solutions.
To achieve this, the researchers behind the glove created a silicone compound holding 44 embedded stretch sensors, combining it with a soft fabric layer. The input device requires very little training, and uses a special constructed set of algorithms to process the sensor data coming through from the gloved hand.
This combination of hardware and software means highly accurate hand and finger movements can be captured in real time. No external cameras or sensors are needed to record the movements, and the gloves themselves can be manufactured at a low cost. They're apparently lightweight and comfortable to wear too.
"This is an already well-studied problem but we found new ways to address it in terms of the sensors employed in our design and our data-driven model," Oliver Glauser from ETH Zurich in Switzerland said in a press statement.
"What is also exciting about this work is the multidisciplinary nature of working on this problem. It required expertise from various fields, including material science, fabrication, electrical engineering, computer graphics, and machine learning."
The glove is still able to capture movements even when the hand is holding something, the researchers say, and because of the configuration of its sensors, it can work in all kinds of lighting conditions too – with no cameras involved, the glove doesn't need to be well lit or within a particular line of sight.
Aside from giving you more accuracy on your next virtual reality adventure, the glove could have applications in robotics, the biomedical industry, and augmented reality (where real world scenes are augmented with computer graphics).
The research team compared its effort against two commercially available gloves, the Manus VR and the CyberGlove II, finding that during testing, the new glove returned the lowest error rate of the three in all but one pose.
Plenty of other companies and research groups are busy working on the smart glove concept, not only to accurately convert real world movements into the digital equivalent, but also to transfer haptic feedback from the virtual world back to real flesh and blood – so you can actually reach out and touch something in VR.
And this new invention will most likely need to work in conjunction with other technology to provide a complete solution. For now at least, it cannot track a hand in 3D space, only the movements of the fingers – so while it does know you're pointing, it doesn't know which direction you're pointing in.
Getting such gloves into mass production continues to be a challenge, but by hitting three key targets – real-time monitoring, standalone operation, and the ability to work in multiple environments – the researchers have made a glove ready for the AR/VR experiences of the future.
The research team will demonstrate the technology at SIGGRAPH 2019 in Los Angeles from July 28. A paper on the development is available online. You can see a video of the glove in action below.