University of Sussex
Browse
QuaLam11.pdf (395.92 kB)

Learning multi-view neighborhood preserving projections

Download (395.92 kB)
conference contribution
posted on 2023-06-08, 16:45 authored by Novi QuadriantoNovi Quadrianto, Christoph Lampert
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 prerequisite 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.

History

Publication status

  • Published

File Version

  • Accepted version

Journal

Proceedings of the 28 th International Conference on Machine Learning; Washington, USA; 28 June - 2 July 2011

Publisher

Association for Computing Machinery

Page range

425-432

Place of publication

New York

ISBN

9781450306195

Department affiliated with

  • Informatics Publications

Full text available

  • Yes

Peer reviewed?

  • Yes

Editors

Tobias Scheffer, Lise Getoor

Legacy Posted Date

2014-02-24

First Open Access (FOA) Date

2017-06-16

First Compliant Deposit (FCD) Date

2017-06-16

Usage metrics

    University of Sussex (Publications)

    Categories

    No categories selected

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC