Overview
The Toshiba CAD model point clouds dataset consists of 12 shape classes. These are bearing1, bearing1_black, block1, bracket1, cog1, cog2, flange1, knob1, mini1, pipe1, piston1, and piston2. Each shape class contains 3D scans of a physical object at different poses. The physical object for each shape class was 3D printed from a known CAD model, as depicted in table 1.
Table 1: CAD models and physical objects of the 12 shape classes
Shape class  CAD model  Physical object 
bearing1  
bearing1_black  
block1  
bracket1  
cog1  
cog2  
flange1  
knob1  
mini1  
pipe1  
piston1  
piston2 
Dataset generation
The geometry of each physical object was scanned 20 times using the realtime 3D reconstruction approach of [3], with a different pose and a different viewing angle at each time. This process generated 20 shape instances for each shape class, in the form of point clouds. Along with the class label, every shape instance has an associated ground truth pose, computed by first approximately registering the relevant CAD model to the point cloud manually, then using the Iterative Closest Point algorithm [1] to refine the registration.Given a test point cloud and set of training point clouds (with known class and pose), the computation of input pose votes poses is a two stage process. In the first stage, local shape features, consisting of a descriptor and a scale, translation and rotation relative to the object, are computed on all the point clouds. This is done by first converting a point cloud to a 128by128by128 voxel volume using a Gaussian on the distance of each voxel centre to the nearest point. Then interest points are localized in the volume across 3D location and scale using the Difference of Gaussians operator, and a canonical orientation for each interest point is computed [2], to generate a local feature pose. Finally a basic, 31dimensional descriptor is computed by simply sampling the volume (at the correct scale) at 31 regularly distributed locations around the interest point.
In the second stage each test feature is matched to the 20 nearest training features, in terms of Euclidean distance between descriptors. Each of these matches generates a vote X_{i}=AB^{1}C, for the test object's pose, A, B and C being the test feature, training feature and training object's ground truth pose respectively. In addition each vote has a weight, λ_{i}, computed as (N_{C}N_{I})^{1}, N_{C} being the number of training instances in the class and N_{I} the number of features found in the feature's particular instance.
The point clouds, the features, the votes, and additionally the lists of rotational symmetries for all shape classes are all available at the Downloads page.
References
[1] P. Besl and N. McKay. A method for registration of 3D shapes. TPAMI, 14(2), 1992.
[2] F. Tombari, S. Salti, and L. Di Stefano. Unique signatures of histograms for local surface description. In ECCV, 2010.
[2] G. Vogiatzis and C. Hernández. Videobased, realtime multi view stereo. Image and Vision Computing, 29(7):434–441, 2011.
Papers
The dataset was introduced in the following paper in a context of 3D object recognition using a votebased approach:A New
Distance for ScaleInvariant 3D Shape Recognition and Registration MinhTri Pham, Oliver J. Woodford, Frank Perbet, Atsuto Maki, Björn Stenger, Roberto Cipolla Cambridge Research Laboratory and University of Cambridge Published in ICCV 2011 [paper, 3.3MB] [project page] 
To our knowledge, it was also used in the following papers:
Distances and Means of Direct Similarities MinhTri Pham, Oliver J. Woodford, Frank Perbet, Atsuto Maki, Riccardo Gherardi, Björn Stenger Cambridge Research Laboratory Published in IJCV 2014 [paper] 

FullAngle Quaternion for Robustly Matching Vectors of 3D Rotations Stephan Liwicki, MinhTri Pham, Stefanos Zafeiriou, Maja Pantic, Björn Stenger Cambridge Research Laboratory Published in CVPR 2014 [paper] 

Demisting the Hough Transform for
3D Shape Recognition and Registration Oliver J. Woodford, MinhTri Pham, Atsuto Maki, Frank Perbet, Björn Stenger Cambridge Research Laboratory Published in BMCV 2011 and IJCV 2013 [paper] 

An
Evaluation of Volumetric Interest Points TszHo Yu, Oliver J. Woodford, Roberto Cipolla Cambridge Research Laboratory and University of Cambridge Published in 3DIM/3DPVT 2011 [paper, 6.3MB] 