Efficient Learning of Domain-Invariant Image Representations

TitleEfficient Learning of Domain-Invariant Image Representations
Publication TypeConference Paper
Year of Publication2013
AuthorsHoffman, J., Rodner E., Donahue J., Darrell T., & Saenko K.
Other Numbers3439
Abstract

We present an algorithm that learns representations which explicitly compensate for domain mismatch and which can be efficiently realized as linear classifiers. Specifically, we form a linear transformation that maps features from the target (test) domain to the source (training) domain as part of training the classifier. We optimize both the transformation and classifier parameters jointly, and introduce an efficient cost function based on misclassification loss.Our method combines several features previously unavailable in a single algorithm: multi-class adaptation through representation learning, ability to map across heterogeneous feature spaces, and scalability to large datasets. We present experiments on several image datasets that demonstrate improved accuracy and computational advantages compared to previous approaches.

Acknowledgment

This work was partially funded by the Deutscher Akademischer Austausch Dienst (DAAD) through a postdoctoral fellowship.

URLhttps://www.icsi.berkeley.edu/pubs/vision/efficientlearning13.pdf
Bibliographic Notes

Proceedings of International Conference on Learning Representations (ICLR 2013), Scottsdale, Arizona

Abbreviated Authors

J. Hoffman, E. Rodner, J. Donahue, T. Darrell, and K. Saenko

ICSI Research Group

Vision

ICSI Publication Type

Article in conference proceedings