Wearable devices with AR/VR applications are regarded as the next generation consumer electronics that might replace smartphones. Currently, the technology is more applied in industrial or professional areas as the devices are still relatively bulky and expensive as they require complex integration of cameras, image processing solutions, power and more.
To tackle the issue, Nvidia developed a new technique for gaze tracking solution, the fundamental technology for VR wearable devices as agreed by many leading industry players, reported VentureBeat. Researchers at Nvidia used LED sensors to track eye, creating a low power gaze tracking solution for minimizing VR hard ware while still achieving good accuracy and low latency.
The LED-based eye-tracking system is based on a ring of infrared LEDs that emit and sense light to determine eye’s position. Nvidia used nine or six LEDs for one eye in two prototypes of the system. These LED-based prototypes only need to integrated with a smartphone and two microcontroller to realize smaller and low cost gaze tracking system for head mounted devices.
Replacing cameras with LEDs for gaze tracking can save power consumption and the space for image processing system to achieve light weight and low cost wearable devices.
However, Nvidia also noted that the system might not be suitable for applications which require very high accuracy such as psychological, neurological or reading researches as the LED system has a median angular error of 0.7 degrees and a mean angular error of 1.1 degrees. Camera-based solutions, on the other hand, deliver “very high accuracy” of error under 0.5 degree.