dc.identifier.uri |
http://dx.doi.org/10.15488/3785 |
|
dc.identifier.uri |
https://www.repo.uni-hannover.de/handle/123456789/3819 |
|
dc.contributor.author |
Menze, Moritz
|
|
dc.contributor.author |
Heipke, Christian
|
|
dc.contributor.author |
Geiger, Andreas
|
|
dc.contributor.editor |
Gall, Juergen
|
|
dc.contributor.editor |
Gehler, Peter
|
|
dc.contributor.editor |
Leibe, Bastian
|
|
dc.date.accessioned |
2018-10-10T08:42:36Z |
|
dc.date.available |
2018-10-10T08:42:36Z |
|
dc.date.issued |
2015 |
|
dc.identifier.citation |
Menze, M.; Heipke, C.; Geiger, A.: Discrete optimization for optical flow. In: Gall, J.; Gehler, P.; Leibe, B. (Eds.): Pattern Recognition. Heidelberg : Springer Verlag, 2015 (Lecture Notes in Computer Science ; 9358), S. 16-28. DOI: https://doi.org/10.1007/978-3-319-24947-6_2 |
|
dc.description.abstract |
We propose to look at large-displacement optical flow from a discrete point of view. Motivated by the observation that sub-pixel accuracy is easily obtained given pixel-accurate optical flow, we conjecture that computing the integral part is the hardest piece of the problem. Consequently, we formulate optical flow estimation as a discrete inference problem in a conditional random field, followed by sub-pixel refinement. Naive discretization of the 2D flow space, however, is intractable due to the resulting size of the label set. In this paper, we therefore investigate three different strategies, each able to reduce computation and memory demands by several orders of magnitude. Their combination allows us to estimate large-displacement optical flow both accurately and efficiently and demonstrates the potential of discrete optimization for optical flow. We obtain state-of-the-art performance on MPI Sintel and KITTI. |
eng |
dc.language.iso |
eng |
|
dc.publisher |
Heidelberg : Springer Verlag |
|
dc.relation.ispartof |
Pattern Recognition |
eng |
dc.relation.ispartofseries |
Lecture Notes in Computer Science ; 9358 |
|
dc.rights |
CC BY-NC 2.5 Unported |
|
dc.rights.uri |
https://creativecommons.org/licenses/by-nc/2.5/ |
|
dc.subject |
Image registration |
eng |
dc.subject |
Optimization |
eng |
dc.subject |
Pattern recognition |
eng |
dc.subject |
Pixels |
eng |
dc.subject |
Conditional random field |
eng |
dc.subject |
Discrete optimization |
eng |
dc.subject |
Inference problem |
eng |
dc.subject |
Large displacements |
eng |
dc.subject |
Optical flow estimation |
eng |
dc.subject |
Orders of magnitude |
eng |
dc.subject |
State-of-the-art performance |
eng |
dc.subject |
Subpixel accuracy |
eng |
dc.subject |
Optical flows |
eng |
dc.subject.classification |
Konferenzschrift |
ger |
dc.subject.ddc |
004 | Informatik
|
ger |
dc.title |
Discrete optimization for optical flow |
eng |
dc.type |
BookPart |
|
dc.type |
Text |
|
dc.relation.isbn |
978-3-319-24946-9 |
|
dc.relation.isbn |
978-3-319-24947-6 |
|
dc.relation.issn |
03029743 |
|
dc.relation.doi |
https://doi.org/10.1007/978-3-319-24947-6_2 |
|
dc.bibliographicCitation.volume |
9358 |
|
dc.bibliographicCitation.firstPage |
16 |
|
dc.bibliographicCitation.lastPage |
28 |
|
dc.description.version |
publishedVersion |
|
tib.accessRights |
frei zug�nglich |
|