Beyond dataset bias: multi-task unaligned shared knowledge transfer

Authors:
Tatiana Tommasi;Novi Quadrianto;Barbara Caputo;Christoph H. Lampert
Affiliations:
Idiap Research Institute, Martigny, Switzerland,École Polytechnique Fédérale de Lausanne, Switzerland;University of Cambridge, UK;Idiap Research Institute, Martigny, Switzerland;IST Austria (Institute of Science and Technology Austria), Klosterneuburg, Austria
Venue:
ACCV'12 Proceedings of the 11th Asian conference on Computer Vision - Volume Part I
Year:
2012

Citing 12
Cited 0

Modeling the Shape of the Scene: A Holistic Representation of the Spatial Envelope

International Journal of Computer Vision
Regularized multi--task learning

Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining
Learning generative visual models from few training examples: An incremental Bayesian approach tested on 101 object categories

Computer Vision and Image Understanding
Distance Metric Learning for Large Margin Nearest Neighbor Classification

The Journal of Machine Learning Research
Domain adaptation with structural correspondence learning

EMNLP '06 Proceedings of the 2006 Conference on Empirical Methods in Natural Language Processing
A Survey on Transfer Learning

IEEE Transactions on Knowledge and Data Engineering
Adapting visual category models to new domains

ECCV'10 Proceedings of the 11th European conference on Computer vision: Part IV
Unbiased look at dataset bias

CVPR '11 Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition
Learning to share visual appearance for multiclass object detection

CVPR '11 Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition
Multiclass transfer learning from unconstrained priors

ICCV '11 Proceedings of the 2011 International Conference on Computer Vision
Domain adaptation for object recognition: An unsupervised approach

ICCV '11 Proceedings of the 2011 International Conference on Computer Vision
Undoing the damage of dataset bias

ECCV'12 Proceedings of the 12th European conference on Computer Vision - Volume Part I

Quantified Score

Hi-index	0.00

Visualization

Abstract

Many visual datasets are traditionally used to analyze the performance of different learning techniques. The evaluation is usually done within each dataset, therefore it is questionable if such results are a reliable indicator of true generalization ability. We propose here an algorithm to exploit the existing data resources when learning on a new multiclass problem. Our main idea is to identify an image representation that decomposes orthogonally into two subspaces: a part specific to each dataset, and a part generic to, and therefore shared between, all the considered source sets. This allows us to use the generic representation as un-biased reference knowledge for a novel classification task. By casting the method in the multi-view setting, we also make it possible to use different features for different databases. We call the algorithm MUST, Multitask Unaligned Shared knowledge Transfer. Through extensive experiments on five public datasets, we show that MUST consistently improves the cross-datasets generalization performance.