This page is part of an archival collection and is no longer actively maintained.

It may contain outdated information and may not meet current or future WCAG accessibility standards. We provide this content, its subpages, and associated links for historical reference only. If you need assistance, please contact support@cs.washington.edu

RGB-D (Kinect) Object Dataset

RGB-D Object Dataset

 Home  People  Dataset  Results  Software  Demos  Publications

The below table sumarizes the results from the different algorithms. All algorithms are evaluated on the same subset of the RGB-D Object Dataset (i.e. subsampled every 5th video frame). For more details on the experimental setting, see our paper.

Category Recognition
Methods Papers RGB Depth RGB-D
SIFT + Texton + Color Histogram + Spin Images + 3D Bounding Boxes ICRA11A 74.5 64.7 83.8
Sparse Distance Learning ICRA11B 78.6 70.2 85.4
RGB-D Kernel Descriptors IROS11 80.7 80.3 86.5
Hierarchical Matcing Pursuit ISER12 82.4 81.2 87.5

Instance Recognition (leave-sequence-out)
Methods Papers RGB Depth RGB-D
SIFT + Texton + Color Histogram + Spin Images + 3D Bounding Boxes ICRA11A 60.7 46.2 74.8
RGB-D Kernel Descriptors IROS11 90.8 54.7 91.2
Hierarchical Matcing Pursuit ISER12 92.1 51.7 92.8

Object Pose Estimation
Methods Papers Median Pose Error Median Pose Error (C) Median Pose Error (I)
Object Pose Tree (OPTree) AAAI11 62.6 51.5 30.2
OPTree + Hierarchical Matcing Pursuit ISER12 20.0 18.7 18.0