Tabula rasa: Model transfer for object category detection
Authors:
- 1 November 2011
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 15505499,p. 2252-2259
Abstract
Our objective is transfer training of a discriminatively trained object category detector, in order to reduce the number of training images required. To this end we propose three transfer learning formulations where a template learnt previously for other categories is used to regularize the training of a new category. All the formulations result in convex optimization problems. Experiments (on PASCAL VOC) demonstrate significant performance gains by transfer learning from one class to another (e.g. motorbike to bicycle), including one-shot learning, specialization from class to a subordinate class (e.g. from quadruped to horse) and transfer using multiple components. In the case of multiple training samples it is shown that a detection performance approaching that of the state of the art can be achieved with substantially fewer training samples.Keywords
This publication has 18 references indexed in Scilit:
- What helps where – and why? Semantic relatedness for knowledge transfer2010
- Safety in numbers: Learning categories from few examples with multi model knowledge transfer2010
- Comparative object similarity for improved recognition with few or no examples2010
- Multiple kernels for object detection2009
- Attribute and simile classifiers for face verification2009
- A shape-based object class model for knowledge transfer2009
- Learning to detect unseen object classes by between-class attribute transfer2009
- Model adaptation with least-squares SVM for adaptive hand prosthetics2009
- The more you know, the less you learn: from knowledge transfer to one-shot learning of object categories2009
- Transfer learning for image classification with sparse prototype representations2008