Learning to be a depth camera for close-range human capture and interaction
Lists
Only you see this
Abstract
We present a machine learning technique for estimating absolute, per-pixel depth using any conventional monocular 2D camera, with minor hardware modifications. Our approach targets close-range human capture and interaction where dense 3D estimation of hands and faces is desired. We use hybrid classification-regression forests to learn how to map from near infrared intensity images to absolute, metric depth in real-time. We demonstrate a variety of human-computer interaction and capture scenarios. Experiments show an accuracy that outperforms a conventional light fall-off baseline, and is comparable to high-quality consumer depth cameras, but with a dramatically reduced cost, power consumption, and form-factor.
BibTeX
@article{fanello2014learning,
author = {Sean Ryan Fanello and Cem Keskin and Shahram Izadi and Pushmeet Kohli and David Kim and David Sweeney and Antonio Criminisi and Jamie Shotton and Sing Bing Kang and Tim Paek},
title = {{Learning to be a depth camera for close-range human capture and interaction}},
journal = {ACM Trans. Graphics},
year = {2014},
doi = {10.1145/2601097.2601223},
}Loading related content...