FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4285044897> ?p ?o ?g. }

• W4285044897 abstract "State-of-the-art methods for optical flow estimation rely on deep learning, which require complex sequential training schemes to reach optimal performances on real-world data. In this work, we introduce the COMBO deep network that explicitly exploits the brightness constancy (BC) model used in traditional methods. Since BC is an approximate physical model violated in several situations, we propose to train a physically-constrained network complemented with a data-driven network. We introduce a unique and meaningful flow decomposition between the physical prior and the data-driven complement, including an uncertainty quantification of the BC model. We derive a joint training scheme for learning the different components of the decomposition ensuring an optimal cooperation, in a supervised but also in a semi-supervised context. Experiments show that COMBO can improve performances over state-of-the-art supervised networks, e.g. RAFT, reaching state-of-the-art results on several benchmarks. We highlight how COMBO can leverage the BC model and adapt to its limitations. Finally, we show that our semi-supervised method can significantly simplify the training procedure." @default.
• W4285044897 created "2022-07-13" @default.
• W4285044897 creator A5017490804 @default.
• W4285044897 creator A5046244833 @default.
• W4285044897 creator A5047597955 @default.
• W4285044897 date "2022-10-23" @default.
• W4285044897 modified "2023-09-30" @default.
• W4285044897 title "Complementing Brightness Constancy with Deep Networks for Optical Flow Prediction" @default.
• W4285044897 cites W1513100184 @default.
• W4285044897 cites W1867429401 @default.
• W4285044897 cites W1964735120 @default.
• W4285044897 cites W1968326286 @default.
• W4285044897 cites W1977194745 @default.
• W4285044897 cites W2005067064 @default.
• W4285044897 cites W2115579991 @default.
• W4285044897 cites W2131747574 @default.
• W4285044897 cites W2156177012 @default.
• W4285044897 cites W2164299365 @default.
• W4285044897 cites W2492517581 @default.
• W4285044897 cites W2507953016 @default.
• W4285044897 cites W2548527721 @default.
• W4285044897 cites W2560474170 @default.
• W4285044897 cites W2604909019 @default.
• W4285044897 cites W2749028154 @default.
• W4285044897 cites W2894983388 @default.
• W4285044897 cites W2904340070 @default.
• W4285044897 cites W2947917444 @default.
• W4285044897 cites W2962864875 @default.
• W4285044897 cites W2963210183 @default.
• W4285044897 cites W2963219046 @default.
• W4285044897 cites W2963782415 @default.
• W4285044897 cites W2963891416 @default.
• W4285044897 cites W2964123261 @default.
• W4285044897 cites W2964156315 @default.
• W4285044897 cites W2967043539 @default.
• W4285044897 cites W2980116241 @default.
• W4285044897 cites W2995511918 @default.
• W4285044897 cites W3034426027 @default.
• W4285044897 cites W3034896357 @default.
• W4285044897 cites W3034953148 @default.
• W4285044897 cites W3035195755 @default.
• W4285044897 cites W3098269293 @default.
• W4285044897 cites W3100280854 @default.
• W4285044897 cites W3100388886 @default.
• W4285044897 cites W3108086282 @default.
• W4285044897 cites W3109908659 @default.
• W4285044897 cites W3122349497 @default.
• W4285044897 cites W3136484018 @default.
• W4285044897 cites W3137898953 @default.
• W4285044897 cites W3147597595 @default.
• W4285044897 cites W3161879860 @default.
• W4285044897 cites W3175558068 @default.
• W4285044897 cites W3184023119 @default.
• W4285044897 cites W3198231706 @default.
• W4285044897 cites W4230014603 @default.
• W4285044897 cites W4249022109 @default.
• W4285044897 cites W764651262 @default.
• W4285044897 doi "https://doi.org/10.1007/978-3-031-19803-8_8" @default.
• W4285044897 hasPublicationYear "2022" @default.
• W4285044897 type Work @default.
• W4285044897 citedByCount "0" @default.
• W4285044897 crossrefType "proceedings-article" @default.
• W4285044897 hasAuthorship W4285044897A5017490804 @default.
• W4285044897 hasAuthorship W4285044897A5046244833 @default.
• W4285044897 hasAuthorship W4285044897A5047597955 @default.
• W4285044897 hasBestOaLocation W42850448971 @default.
• W4285044897 hasConcept C104317684 @default.
• W4285044897 hasConcept C108583219 @default.
• W4285044897 hasConcept C112313634 @default.
• W4285044897 hasConcept C119857082 @default.
• W4285044897 hasConcept C127716648 @default.
• W4285044897 hasConcept C151730666 @default.
• W4285044897 hasConcept C153083717 @default.
• W4285044897 hasConcept C154945302 @default.
• W4285044897 hasConcept C165696696 @default.
• W4285044897 hasConcept C185592680 @default.
• W4285044897 hasConcept C188082640 @default.
• W4285044897 hasConcept C2779343474 @default.
• W4285044897 hasConcept C38652104 @default.
• W4285044897 hasConcept C41008148 @default.
• W4285044897 hasConcept C55493867 @default.
• W4285044897 hasConcept C86803240 @default.
• W4285044897 hasConceptScore W4285044897C104317684 @default.
• W4285044897 hasConceptScore W4285044897C108583219 @default.
• W4285044897 hasConceptScore W4285044897C112313634 @default.
• W4285044897 hasConceptScore W4285044897C119857082 @default.
• W4285044897 hasConceptScore W4285044897C127716648 @default.
• W4285044897 hasConceptScore W4285044897C151730666 @default.
• W4285044897 hasConceptScore W4285044897C153083717 @default.
• W4285044897 hasConceptScore W4285044897C154945302 @default.
• W4285044897 hasConceptScore W4285044897C165696696 @default.
• W4285044897 hasConceptScore W4285044897C185592680 @default.
• W4285044897 hasConceptScore W4285044897C188082640 @default.
• W4285044897 hasConceptScore W4285044897C2779343474 @default.
• W4285044897 hasConceptScore W4285044897C38652104 @default.
• W4285044897 hasConceptScore W4285044897C41008148 @default.
• W4285044897 hasConceptScore W4285044897C55493867 @default.
• W4285044897 hasConceptScore W4285044897C86803240 @default.
• W4285044897 hasLocation W42850448971 @default.
• W4285044897 hasLocation W42850448972 @default.

• next