Filters








3 Hits in 1.1 sec

HyperDepth: Learning Depth from Structured Light without Matching

Sean Ryan Fanello, Christoph Rhemann, Vladimir Tankovich, Adarsh Kowdle, Sergio Orts Escolano, David Kim, Shahram Izadi
2016 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
Structured light sensors are popular due to their robustness to untextured scenes and multipath.  ...  We contribute an algorithm for solving this correspondence problem efficiently, without compromising depth accuracy.  ...  Learning to Recognize Structured Light In this section we reformulate the spatial structured light correspondence problem from a machine learning perspec-tive, and show how disparity maps with subpixel  ... 
doi:10.1109/cvpr.2016.587 dblp:conf/cvpr/FanelloRTKOKI16 fatcat:semu6akx6jb2veqw2zvpevz5k4

UltraStereo: Efficient Learning-Based Matching for Active Stereo Systems

Sean Ryan Fanello, Julien Valentin, Christoph Rhemann, Adarsh Kowdle, Vladimir Tankovich, Philip Davidson, Shahram Izadi
2017 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
compared to time of flight or traditional structured light techniques.  ...  We efficiently solve the specialized problem of stereo matching under active illumination using a new learning-based algorithm.  ...  [14] use diffuse infrared light to learn a shape from shading mapping from IR intensity to depth, but the technique works only for hands and faces in a very limited range.  ... 
doi:10.1109/cvpr.2017.692 dblp:conf/cvpr/FanelloVRKTDI17 fatcat:auf6kvbp5nfabi2glqbcovpiva

Connecting the Dots: Learning Representations for Active Monocular Depth Estimation

Gernot Riegler, Yiyi Liao, Simon Donne, Vladlen Koltun, Andreas Geiger
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
Further, we demonstrate that the projected pattern of the structured light sensor can be reliably separated from the ambient information.  ...  We propose a technique for depth estimation with a monocular structured-light camera, i.e., a calibrated stereo set-up with one camera and one laser projector.  ...  [41] exploit ideas from self-supervised learning to train an active stereo network without needing ground-truth depth.  ... 
doi:10.1109/cvpr.2019.00781 dblp:conf/cvpr/RieglerLDKG19 fatcat:p6gx6ir4b5behegpt55xz3qhwa