A shape-based object class model for knowledge transfer
Michael M. Stark, Michael Goesele, Bernt Schiele · IEEE International Conference on Computer Vision (ICCV) · 2009
Object class models trained on hundreds or thousands of images have shown to enable robust detection. Transferring knowledge from such models to new object classes trained from a few or even as little as one training instance however is still in its infancy. This paper designs a shape-based model that allows to easily and explicitly transfer knowledge on three different levels: transfer of individual parts' shape and appearance information, transfer of local symmetry between parts, and transfer of part topology. Due to the factorized form of the model, knowledge can either be transferred for the complete model or just partial knowledge corresponding to certain aspects of the model. The experiments clearly demonstrate that the proposed model is competitive with the state-of-the-art and enables both full and partial knowledge transfer.
Cited by 82 works. doi:10.1109/iccv.2009.5459231
Authors
References
- Object class recognition by unsupervised scale-invariant learning
- Markov Chain Monte Carlo in Practice
- One-shot learning of object categories
- Learning to detect natural image boundaries using local brightness, color, and texture cues
- Description and recognition of curved objects☆
- Markov Chain Monte Carlo in Practice
- Exploiting Object Hierarchy: Combining Models from Different Category Levels
- From Aardvark to Zorro: A Benchmark for Mammal Image Classification
- Sharing Visual Features for Multiclass and Multiview Object Detection
- Image Parsing: Unifying Segmentation, Detection, and Recognition