Devon: deformable volume network for learning optical flow

Yao Lu, Jack Valmadre, Heng Wang, Juho Kannala, Mehrtash Harandi, Philip H.S. Torr

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

Abstract

We propose a new neural network module, Deformable Cost Volume, for learning large displacement optical flow. The module does not distort the original images or their feature maps and therefore avoids the artifacts associated with warping. Based on this module, a new neural network model is proposed. The full version of this paper can be found online (https://arxiv.org/abs/1802.07351 ).

Original languageEnglish
Title of host publicationComputer Vision – ECCV 2018 Workshops
Subtitle of host publicationMunich, Germany, September 8–14, 2018 Proceedings, Part VI
EditorsLaura Leal-Taixé, Stefan Roth
Place of PublicationCham Switzerland
PublisherSpringer
Pages673-677
Number of pages5
ISBN (Electronic)9783030110246
ISBN (Print)9783030110239
DOIs
Publication statusPublished - 2019
EventWhat Is Optical Flow for? Workshop 2018 - Munich, Germany
Duration: 14 Sep 201814 Sep 2018

Publication series

NameLecture Notes in Computer Science
PublisherSpringer
Volume11134
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Workshop

WorkshopWhat Is Optical Flow for? Workshop 2018
CountryGermany
CityMunich
Period14/09/1814/09/18

Cite this

Lu, Y., Valmadre, J., Wang, H., Kannala, J., Harandi, M., & Torr, P. H. S. (2019). Devon: deformable volume network for learning optical flow. In L. Leal-Taixé, & S. Roth (Eds.), Computer Vision – ECCV 2018 Workshops: Munich, Germany, September 8–14, 2018 Proceedings, Part VI (pp. 673-677). (Lecture Notes in Computer Science; Vol. 11134). Springer. https://doi.org/10.1007/978-3-030-11024-6_50