Microsoft is working on a mixed reality headset that uses artificial intelligence (AI) to boost processing speed.
The second version of the processing unit for Microsoft’s HoloLens will employ AI to work on some of the intensive processing work required for mixed reality.
Director of Science at HoloLens, Marc Pollefeys, argues that deep learning has massively improved over the past few years, describing progress in recognition problems as ‘astonishing.’
Pollefeys attributes this progress to Deep Neural Networks (DNNs), but also notes that DNNs present two challenges: ‘They require large amounts of labelled data for training, and they require a type of compute that is not amenable to current general-purpose processor/memory architectures.’
Most efforts at working this out so far have been based on ‘enhancing existing cloud computing fabrics,’ but Pollefeys states that HoloLens is ‘in the business of making untethered mixed reality devices.’
To have this completely independent device, in which a custom hologram processing unit processes all information, including Microsoft’s custom time-of-flight depth sensor, head-tracking cameras, inertial measurement unit (IMU) and infrared camera, Microsoft has incorporated an AI processor which will ‘natively and flexibly implement DNNs.’
This technology may have wider implications for certain types of cloud services. As such devices require instantaneous processing, analysts think that AI could become significantly more prevalent, meaning external cloud processing may take a hit. Jim McGregor, an analyst at Tirias Research, believes that by 2025, ‘every device people interact with will have AI built in.’
There are significant real-life implications which mean processing has to be instant, according to McGregor. ‘The consumer is going to expect to have almost no lag and to do real-time processing.’
‘For an autonomous car, you can’t afford the time to send it back to the cloud to make the decisions to avoid the crash, to avoid hitting a person. The amount of data coming out of autonomous vehicles is tremendous; you can’t send all of that to the cloud,’ he said.