Hybrid Bayesian Eigenobjects: Combining Linear Subspace and Deep Network Methods for 3D Robot Vision

We introduce Hybrid Bayesian Eigenobjects (HBEOs), a novel representation for 3D objects designed to allow a robot to jointly estimate the pose, class, and full 3D geometry of a novel object observed from a single viewpoint in a single practical framework. By combining both linear subspace methods and deep convolutional prediction, HBEOs efficiently learn nonlinear object representations without directly regressing into high-dimensional space. HBEOs also remove the onerous and generally impractical necessity of input data voxelization prior to inference. We experimentally evaluate the suitability of HBEOs to the challenging task of joint pose, class, and shape inference on novel objects and show that, compared to preceding work, HBEOs offer dramatically improved performance in all three tasks along with several orders of magnitude faster runtime performance.

Authors

• Benjamin Burchfiel
• George Konidaris 0001

Keywords

• Three-dimensional displays
• Robots
• Bayes methods
• Pose estimation
• Databases
• Principal component analysis
• Task analysis
• Deep Network
• Linear Subspace
• High-dimensional
• Linear Method
• Object Classification
• 3D Geometry
• Object Pose
• Object Geometry
• Runtime Performance
• Degrees Of Freedom
• Convolutional Network
• Feature Learning
• Point Cloud
• Network Output
• Deep Convolutional Network
• Depth Images
• Objective Space
• 3D Classification
• Robotic Tasks
• Position In The World
• Low-dimensional Subspace
• Complete Object
• Human Pose Estimation
• Subspace Projection
• Ground-truth Class
• Subspace Learning
• 3D Joint
• Shape Completion
• Gaussian Mixture Model