Quadrianto, Novi; Lampert , ChristophISTA
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 expresses 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.
425 - 432
ICML: International Conference on Machine Learning
2011-06-28 – 2011-07-02
Quadrianto N, Lampert C. Learning multi-view neighborhood preserving projections. In: Omnipress; 2011:425-432.
Quadrianto, N., & Lampert, C. (2011). Learning multi-view neighborhood preserving projections (pp. 425–432). Presented at the ICML: International Conference on Machine Learning, Bellevue, USA: Omnipress.
Quadrianto, Novi, and Christoph Lampert. “Learning Multi-View Neighborhood Preserving Projections,” 425–32. Omnipress, 2011.
N. Quadrianto and C. Lampert, “Learning multi-view neighborhood preserving projections,” presented at the ICML: International Conference on Machine Learning, Bellevue, USA, 2011, pp. 425–432.
Quadrianto N, Lampert C. 2011. Learning multi-view neighborhood preserving projections. ICML: International Conference on Machine Learning, 425–432.
Quadrianto, Novi, and Christoph Lampert. Learning Multi-View Neighborhood Preserving Projections. Omnipress, 2011, pp. 425–32.