Constructing implicit 3D shape models for pose estimation
Authors:
- 1 September 2009
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 15505499,p. 1341-1348
Abstract
We present a system that constructs “implicit shape models” for classes of rigid 3D objects and utilizes these models to estimating the pose of class instances in single 2D images. We use the framework of implicit shape models to construct a voting procedure that allows for 3D transformations and projection and accounts for self occlusion. The model is comprised of a collection of learned features, their 3D locations, their appearances in different views, and the set of views in which they are visible. We further learn the parameters of a model from training images by applying a method that relies on factorization. We demonstrate the utility of the constructed models by applying them in pose estimation experiments to recover the viewpoint of class instances.Keywords
This publication has 20 references indexed in Scilit:
- Pose estimation for category specific multiview object localization2009
- Factorization for non-rigid and articulated structure using metric projections2009
- Linear embeddings in non-rigid structure from motion2009
- Viewpoint-independent object class detection using 3D Feature Maps2008
- An Exemplar Model for Learning Object Classes2007
- Flexible Object Models for Category-Level 3D Object Recognition2007
- 3D generic object categorization, localization and pose estimation2007
- Distinctive Image Features from Scale-Invariant KeypointsInternational Journal of Computer Vision, 2004
- Linear Fitting with Missing Data for Structure-from-MotionComputer Vision and Image Understanding, 2001
- Recognizing solid objects by alignment with an imageInternational Journal of Computer Vision, 1990