Descriptor learning using convex optimisation

Conference item

The objective of this work is to learn descriptors suitable for the sparse feature detectors used in viewpoint invariant matching. We make a number of novel contributions towards this goal: first, it is shown that learning the pooling regions for the descriptor can be formulated as a convex optimisation problem selecting the regions using sparsity; second, it is shown that dimensionality reduction can also be formulated as a convex optimisation problem, using the nuclear norm to reduce dimensionality. Both of these problems use large margin discriminative learning methods. The third contribution is a new method of obtaining the positive and negative training data in a weakly supervised manner. And, finally, we employ a state-of-the-art stochastic optimizer that is efficient and well matched to the non-smooth cost functions proposed here. It is demonstrated that the new learning methods improve over the state of the art in descriptor learning for large scale matching, Brown et al. [2], and large scale object retrieval, Philbin et al. [10].

Authors: Simonyan, K; Vedaldi, A; Zisserman, A

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: Springer
• Host title: Computer Vision – ECCV 2012: 12th European Conference on Computer Vision, Florence, Italy, October 7-13, 2012, Proceedings, Part I
• Pages: 243-256
• Series: Lecture Notes in Computer Science
• Series number: 7572
• Place of publication: Berlin / Heidelberg
• Publication date: 2012-09-26
• Acceptance date: 2012-06-25
• Event title: 12th European Conference on Computer Vision (ECCV 2012)
• Event location: Florence, Italy
• Event website: https://eccv2012.unifi.it/
• Event start date: 2012-10-07
• Event end date: 2012-10-13
• DOI: 10.1007/978-3-642-33718-5_18
• EISSN: 1611-3349
• ISSN: 0302-9743
• EISBN: 9783642337185
• ISBN: 9783642337178
• Language: English
• Keywords: feature pair; image retrieval; convex optimisation problem; pooling region; convex optimisation