Learning multi-view neighborhood preserving projections Conference Paper


Author(s): Quadrianto, Novi; Lampert, Christoph
Title: Learning multi-view neighborhood preserving projections
Affiliation IST Austria
Abstract: We address the problem of metric learning for multi-view data, namely the construction of embedding projections from data in different representations into a shared feature space, such that the Euclidean distance in this space provides a meaningful within-view as well as between-view similarity. Our motivation stems from the problem of cross-media retrieval tasks, where the availability of a joint Euclidean distance function is a pre-requisite to allow fast, in particular hashing-based, nearest neighbor queries. We formulate an objective function that ex- presses the intuitive concept that matching samples are mapped closely together in the output space, whereas non-matching samples are pushed apart, no matter in which view they are available. The resulting optimization problem is not convex, but it can be decomposed explicitly into a convex and a concave part, thereby allowing efficient optimization using the convex-concave procedure. Experiments on an image retrieval task show that nearest-neighbor based cross- view retrieval is indeed possible, and the proposed technique improves the retrieval accuracy over baseline techniques.
Conference Title: ICML: International Conference on Machine Learning
Conference Dates: June 28 - July 2, 2011
Conference Location: Bellevue, Washington, USA
Publisher: Omnipress  
Date Published: 2011-01-01
Start Page: 425
End Page: 432
URL:
Open access: no
IST Austria Authors
  1. Christoph Lampert
    87 Lampert
Related IST Austria Work