File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Conference Paper: S2F: Slow-to-fast interpolator flow

TitleS2F: Slow-to-fast interpolator flow
Authors
Issue Date2017
Citation
Proceedings - 30th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2017, 2017, v. 2017-January, p. 3767-3776 How to Cite?
AbstractWe introduce a method to compute optical flow at multiple scales of motion, without resorting to multiresolution or combinatorial methods. It addresses the key problem of small objects moving fast, and resolves the artificial binding between how large an object is and how fast it can move before being diffused away by classical scale-space. Even with no learning, it achieves top performance on the most challenging optical flow benchmark. Moreover, the results are interpretable, and indeed we list the assumptions underlying our method explicitly. The key to our approach is the matching progression from slow to fast, as well as the choice of interpolation method, or equivalently the prior, to fill in regions where the data allows it. We use several offthe-shelf components, with relatively low sensitivity to parameter tuning. Computational cost is comparable to the state-of-the-art.
Persistent Identifierhttp://hdl.handle.net/10722/325377
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorYang, Yanchao-
dc.contributor.authorSoatto, Stefano-
dc.date.accessioned2023-02-27T07:32:22Z-
dc.date.available2023-02-27T07:32:22Z-
dc.date.issued2017-
dc.identifier.citationProceedings - 30th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2017, 2017, v. 2017-January, p. 3767-3776-
dc.identifier.urihttp://hdl.handle.net/10722/325377-
dc.description.abstractWe introduce a method to compute optical flow at multiple scales of motion, without resorting to multiresolution or combinatorial methods. It addresses the key problem of small objects moving fast, and resolves the artificial binding between how large an object is and how fast it can move before being diffused away by classical scale-space. Even with no learning, it achieves top performance on the most challenging optical flow benchmark. Moreover, the results are interpretable, and indeed we list the assumptions underlying our method explicitly. The key to our approach is the matching progression from slow to fast, as well as the choice of interpolation method, or equivalently the prior, to fill in regions where the data allows it. We use several offthe-shelf components, with relatively low sensitivity to parameter tuning. Computational cost is comparable to the state-of-the-art.-
dc.languageeng-
dc.relation.ispartofProceedings - 30th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2017-
dc.titleS2F: Slow-to-fast interpolator flow-
dc.typeConference_Paper-
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.doi10.1109/CVPR.2017.401-
dc.identifier.scopuseid_2-s2.0-85041902133-
dc.identifier.volume2017-January-
dc.identifier.spage3767-
dc.identifier.epage3776-
dc.identifier.isiWOS:000418371403090-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats