Sublabel-accurate convex relaxation of vectorial multilabel energies
Laude E, Moellenhoff T, Moeller M, Lellmann J, Cremers D (2016)
Publication Type: Conference contribution
Publication year: 2016
Journal
Publisher: Springer Verlag
Book Volume: 9905 LNCS
Pages Range: 614-627
Conference Proceedings Title: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Event location: Amsterdam, NLD
ISBN: 9783319464473
DOI: 10.1007/978-3-319-46448-0_37
Abstract
Convex relaxations of multilabel problems have been demonstrated to produce provably optimal or near-optimal solutions to a variety of computer vision problems. Yet, they are of limited practical use as they require a fine discretization of the label space, entailing a huge demand in memory and runtime. In this work, we propose the first sublabel accurate convex relaxation for vectorial multilabel problems. Our key idea is to approximate the dataterm in a piecewise convex (rather than piecewise linear) manner. As a result we have a more faithful approximation of the original cost function that provides a meaningful interpretation for fractional solutions of the relaxed convex problem.
Involved external institutions
Germany (DE)How to cite
APA:
Laude, E., Moellenhoff, T., Moeller, M., Lellmann, J., & Cremers, D. (2016). Sublabel-accurate convex relaxation of vectorial multilabel energies. In Bastian Leibe, Jiri Matas, Nicu Sebe, Max Welling (Eds.), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (pp. 614-627). Amsterdam, NLD: Springer Verlag.
MLA:
Laude, Emanuel, et al. "Sublabel-accurate convex relaxation of vectorial multilabel energies." Proceedings of the 14th European Conference on Computer Vision, ECCV 2016, Amsterdam, NLD Ed. Bastian Leibe, Jiri Matas, Nicu Sebe, Max Welling, Springer Verlag, 2016. 614-627.
BibTeX: Download