Unsupervised feature learning for 3D scene labeling

This paper presents an approach for labeling objects in 3D scenes. We introduce HMP3D, a hierarchical sparse coding technique for learning features from 3D point cloud data. HMP3D classifiers are trained using a synthetic dataset of virtual scenes generated using CAD models from an online database. Our scene labeling system combines features learned from raw RGB-D images and 3D point clouds directly, without any hand-designed features, to assign an object label to every 3D point in the scene. Experiments on the RGB-D Scenes Dataset v. 2 demonstrate that the proposed approach can be used to label indoor scenes containing both small tabletop objects and large furniture pieces.

Authors

• Kevin Lai 0001
• Liefeng Bo
• Dieter Fox

Keywords

• Three-dimensional displays
• Labeling
• Feature extraction
• Videos
• Matching pursuit algorithms
• Dictionaries
• Solid modeling
• Feature Learning
• 3D Scene
• Unsupervised Feature Learning
• Online Database
• Point Cloud
• 3D Data
• 3D Point
• 3D Point Cloud
• Objects In The Scene
• CAD Model
• Sparse Coding
• Hierarchical Technique
• RGB-D Images
• Scene Point
• Sparse Techniques
• Hierarchical Coding
• 3D Point Cloud Data
• Support Vector Machine
• Feature Representation
• K-nearest Neighbor
• Spatial Pooling
• Pixel In Frame
• Coffee Table
• Markov Random Field
• RGB Images
• Spatial Pyramid Pooling
• Surface Normals
• Depth Images
• Codeword
• Feature Representation Learning