Deep self-correlation descriptor for dense cross-modal correspondence

Seungryong Kim, Dongbo Min, Stephen Lin, Kwanghoon Sohn

Research output: Chapter in Book/Report/Conference proceedingConference contribution

7 Citations (Scopus)

Abstract

We present a novel descriptor, called deep self-correlation (DSC), designed for establishing dense correspondences between images taken under different imaging modalities, such as different spectral ranges or lighting conditions. Motivated by local self-similarity (LSS), we formulate a novel descriptor by leveraging LSS in a deep architecture, leading to better discriminative power and greater robustness to non-rigid image deformations than state-of-the-art descriptors. The DSC first computes self-correlation surfaces over a local support window for randomly sampled patches, and then builds hierarchical self-correlation surfaces by performing an average pooling within a deep architecture. Finally, the feature responses on the self-correlation surfaces are encoded through a spatial pyramid pooling in a circular configuration. In contrast to convolutional neural networks (CNNs) based descriptors, the DSC is trainingfree, is robust to cross-modal imaging, and can be densely computed in an efficient manner that significantly reduces computational redundancy. The state-of-the-art performance of DSC on challenging cases of cross-modal image pairs is demonstrated through extensive experiments.

Original languageEnglish
Title of host publicationComputer Vision - 14th European Conference, ECCV 2016, Proceedings
EditorsBastian Leibe, Jiri Matas, Max Welling, Nicu Sebe
PublisherSpringer Verlag
Pages679-695
Number of pages17
ISBN (Print)9783319464831
DOIs
Publication statusPublished - 2016 Jan 1

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9912 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Deep self-correlation descriptor for dense cross-modal correspondence'. Together they form a unique fingerprint.

  • Cite this

    Kim, S., Min, D., Lin, S., & Sohn, K. (2016). Deep self-correlation descriptor for dense cross-modal correspondence. In B. Leibe, J. Matas, M. Welling, & N. Sebe (Eds.), Computer Vision - 14th European Conference, ECCV 2016, Proceedings (pp. 679-695). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9912 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-46484-8_41