Recognition of 3D Objects by Learning from Correspondences in a Sequence of Unlabeled Training Images.
Raimund LeitnerHorst BischofPublished in: DAGM-Symposium (2005)
Keyphrases
- d objects
- object recognition
- polyhedral objects
- invariant recognition
- viewpoint
- active learning
- supervised learning
- pose estimation
- recognition and pose estimation
- three dimensional
- machine learning
- recognizing objects
- occluded objects
- image features
- multiple views
- point correspondences
- line drawings
- object representation
- action sequences
- image data
- computer vision