Hello,
I know the Microsoft Kinect SDK has a human skeleton pose detection algorithm. My question is the following:
Is there something equivalent for the new Microsoft Hololense? Like the Kinect, it generates scene depth maps so in theory, it has the data needed to do the work but I also know the processing power of the Hololense is very limited (cpu is an Atom processor) compared to the high-end PC to which the Kinect would ne connected.
As the Hololense subject is very new, any link or ideas about how to concretely implement human pose estimation with it would be welcome!
Thanks,
Bruno Martin
https://www.microsoft.com/microsoft-hololens/en-us
https://developer.microsoft.com/en-us/windows/kinect/develop
Interesting news combining Kinect with an Hololens:
http://hololens.reality.news/how-to/add-full-body-motion-tracking-hololens-with-kinect-0174419/
These are good papers. Do you intend to publish the code under a commercially usable license?
@Meng Ding,i think the Hololens uses the two depth sensor instead of two rgb sensors.
1st,getting a density point cloud.The feature based or optical-flow based is not enough to generate such density mesh as Hololens does now.
2nd,rgb sensor is not effective in dark environment,but Hololens does.RGBD sensor is more robust than rgb sensor.
For these two reasons,i think the MS would choose RGBD sensor.