FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4285066052> ?p ?o ?g. }

• W4285066052 endingPage "20859" @default.
• W4285066052 startingPage "20850" @default.
• W4285066052 abstract "Optical flow estimation is a fundamental problem of computer vision and has many applications in the fields of robot learning and autonomous driving. This paper reveals novel geometric principles of optical flow based on the insight and detailed definition of non-occlusion. Then, two novel loss functions are proposed for the unsupervised learning of optical flow based on the geometric principles of non-occlusion. Specifically, after the occlusion part of the images are masked, the flowing process of pixel coordinate points is carefully considered and geometric constraints are conducted based on the geometric principles of optical flow. First, the optical flow vectors of neighboring pixel coordinate points in the first frame will not intersect during the pixel displacement to the second frame. Secondly, when the cluster containing adjacent four pixel coordinate points in the first frame moves to the second frame, no other pixel coordinate points will flow into the quadrilateral formed by them. According to the two geometrical constraints, the optical flow non-intersection loss and the optical flow non-blocking loss in the non-occlusion regions are proposed. Two loss functions punish the irregular and inexact optical flows in the non-occlusion regions. The experiments on datasets demonstrated that the proposed unsupervised losses of optical flow based on the geometric principles in non-occlusion regions make the estimated optical flow more refined in detail, and improve the performance of unsupervised learning of optical flow. In addition, the experiments training on synthetic data and evaluating on real data show that the generalization ability of optical flow network is improved by our proposed unsupervised approach. Our codes are open sourced at: <uri xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>https://github.com/IRMVLab/NccFlow</uri> ." @default.
• W4285066052 created "2022-07-13" @default.
• W4285066052 creator A5013563796 @default.
• W4285066052 creator A5036208549 @default.
• W4285066052 creator A5074890105 @default.
• W4285066052 date "2022-11-01" @default.
• W4285066052 modified "2023-10-18" @default.
• W4285066052 title "Unsupervised Learning of Optical Flow With Non-Occlusion From Geometry" @default.
• W4285066052 cites W1513100184 @default.
• W4285066052 cites W1578285471 @default.
• W4285066052 cites W1867429401 @default.
• W4285066052 cites W2033959528 @default.
• W4285066052 cites W2133665775 @default.
• W4285066052 cites W2147253850 @default.
• W4285066052 cites W2150066425 @default.
• W4285066052 cites W2548527721 @default.
• W4285066052 cites W2560474170 @default.
• W4285066052 cites W2604909019 @default.
• W4285066052 cites W2770902190 @default.
• W4285066052 cites W2793872863 @default.
• W4285066052 cites W2890949887 @default.
• W4285066052 cites W2894983388 @default.
• W4285066052 cites W2898424515 @default.
• W4285066052 cites W2904340070 @default.
• W4285066052 cites W2949023359 @default.
• W4285066052 cites W2962864875 @default.
• W4285066052 cites W2963782415 @default.
• W4285066052 cites W2963891416 @default.
• W4285066052 cites W2964968086 @default.
• W4285066052 cites W2967043539 @default.
• W4285066052 cites W2982361806 @default.
• W4285066052 cites W3007774126 @default.
• W4285066052 cites W3034308492 @default.
• W4285066052 cites W3034896357 @default.
• W4285066052 cites W3035235001 @default.
• W4285066052 cites W3069056889 @default.
• W4285066052 cites W3108086282 @default.
• W4285066052 cites W3130804049 @default.
• W4285066052 cites W3173486695 @default.
• W4285066052 cites W3182809356 @default.
• W4285066052 cites W4226041473 @default.
• W4285066052 cites W4226491373 @default.
• W4285066052 cites W4226494925 @default.
• W4285066052 cites W4313016775 @default.
• W4285066052 cites W764651262 @default.
• W4285066052 doi "https://doi.org/10.1109/tits.2022.3182858" @default.
• W4285066052 hasPublicationYear "2022" @default.
• W4285066052 type Work @default.
• W4285066052 citedByCount "5" @default.
• W4285066052 countsByYear W42850660522022 @default.
• W4285066052 countsByYear W42850660522023 @default.
• W4285066052 crossrefType "journal-article" @default.
• W4285066052 hasAuthorship W4285066052A5013563796 @default.
• W4285066052 hasAuthorship W4285066052A5036208549 @default.
• W4285066052 hasAuthorship W4285066052A5074890105 @default.
• W4285066052 hasConcept C115961682 @default.
• W4285066052 hasConcept C121332964 @default.
• W4285066052 hasConcept C135628077 @default.
• W4285066052 hasConcept C151756577 @default.
• W4285066052 hasConcept C154945302 @default.
• W4285066052 hasConcept C155542232 @default.
• W4285066052 hasConcept C160633673 @default.
• W4285066052 hasConcept C2524010 @default.
• W4285066052 hasConcept C31972630 @default.
• W4285066052 hasConcept C33923547 @default.
• W4285066052 hasConcept C38349280 @default.
• W4285066052 hasConcept C41008148 @default.
• W4285066052 hasConcept C97355855 @default.
• W4285066052 hasConceptScore W4285066052C115961682 @default.
• W4285066052 hasConceptScore W4285066052C121332964 @default.
• W4285066052 hasConceptScore W4285066052C135628077 @default.
• W4285066052 hasConceptScore W4285066052C151756577 @default.
• W4285066052 hasConceptScore W4285066052C154945302 @default.
• W4285066052 hasConceptScore W4285066052C155542232 @default.
• W4285066052 hasConceptScore W4285066052C160633673 @default.
• W4285066052 hasConceptScore W4285066052C2524010 @default.
• W4285066052 hasConceptScore W4285066052C31972630 @default.
• W4285066052 hasConceptScore W4285066052C33923547 @default.
• W4285066052 hasConceptScore W4285066052C38349280 @default.
• W4285066052 hasConceptScore W4285066052C41008148 @default.
• W4285066052 hasConceptScore W4285066052C97355855 @default.
• W4285066052 hasFunder F4320321001 @default.
• W4285066052 hasFunder F4320321885 @default.
• W4285066052 hasIssue "11" @default.
• W4285066052 hasLocation W42850660521 @default.
• W4285066052 hasOpenAccess W4285066052 @default.
• W4285066052 hasPrimaryLocation W42850660521 @default.
• W4285066052 hasRelatedWork W121273120 @default.
• W4285066052 hasRelatedWork W2123374136 @default.
• W4285066052 hasRelatedWork W2171116555 @default.
• W4285066052 hasRelatedWork W2337415362 @default.
• W4285066052 hasRelatedWork W2740820121 @default.
• W4285066052 hasRelatedWork W2900363363 @default.
• W4285066052 hasRelatedWork W3096301266 @default.
• W4285066052 hasRelatedWork W3098068165 @default.
• W4285066052 hasRelatedWork W4281553171 @default.
• W4285066052 hasRelatedWork W4312857205 @default.
• W4285066052 hasVolume "23" @default.

• next