Matches in SemOpenAlex for { <https://semopenalex.org/work/W3136181114> ?p ?o ?g. }
- W3136181114 abstract "There are increasing interests of studying the structure-from-motion (SfM) problem with machine learning techniques. While earlier methods directly learn a mapping from images to depth maps and camera poses, more recent works enforce multi-view geometry constraints through optimization embedded in the learning framework. This paper presents a novel optimization method based on recurrent neural networks to further exploit the potential of neural networks in SfM. Specifically, our neural optimizer alternately updates the depth and camera poses through iterations to minimize a feature-metric cost, and two gated recurrent units iteratively improve the results by tracing historical information. In this way, our network is a gradient-free zeroth-order optimizer designed for SfM and can be applied to both supervised and self-supervised SfM. Extensive experimental results demonstrate that our method outperforms previous methods and is more efficient in computation and memory consumption than cost-volume-based methods. In particular, our self-supervised method outperforms previous supervised methods on the KITTI and ScanNet datasets. Our source code is available at this https URL." @default.
- W3136181114 created "2021-03-29" @default.
- W3136181114 creator A5013549550 @default.
- W3136181114 creator A5029170925 @default.
- W3136181114 creator A5052772478 @default.
- W3136181114 creator A5060851814 @default.
- W3136181114 creator A5069033376 @default.
- W3136181114 creator A5084953118 @default.
- W3136181114 date "2021-03-24" @default.
- W3136181114 modified "2023-09-26" @default.
- W3136181114 title "DRO: Deep Recurrent Optimizer for Structure-from-Motion." @default.
- W3136181114 cites W1924770834 @default.
- W3136181114 cites W1985238052 @default.
- W3136181114 cites W2021851106 @default.
- W3136181114 cites W2124313187 @default.
- W3136181114 cites W2133665775 @default.
- W3136181114 cites W2171740948 @default.
- W3136181114 cites W2190691619 @default.
- W3136181114 cites W2194775991 @default.
- W3136181114 cites W2198618282 @default.
- W3136181114 cites W2474281075 @default.
- W3136181114 cites W2561074213 @default.
- W3136181114 cites W2593414960 @default.
- W3136181114 cites W2594519801 @default.
- W3136181114 cites W2609883120 @default.
- W3136181114 cites W2624086852 @default.
- W3136181114 cites W2808456995 @default.
- W3136181114 cites W2887825894 @default.
- W3136181114 cites W2889827408 @default.
- W3136181114 cites W2949023359 @default.
- W3136181114 cites W2962793285 @default.
- W3136181114 cites W2962816904 @default.
- W3136181114 cites W2962960377 @default.
- W3136181114 cites W2963216700 @default.
- W3136181114 cites W2963488291 @default.
- W3136181114 cites W2963583471 @default.
- W3136181114 cites W2963706662 @default.
- W3136181114 cites W2963906250 @default.
- W3136181114 cites W2963970238 @default.
- W3136181114 cites W2964314455 @default.
- W3136181114 cites W2964968086 @default.
- W3136181114 cites W2970791107 @default.
- W3136181114 cites W2971000934 @default.
- W3136181114 cites W2981732213 @default.
- W3136181114 cites W2985775862 @default.
- W3136181114 cites W2995124519 @default.
- W3136181114 cites W2997687993 @default.
- W3136181114 cites W3034364596 @default.
- W3136181114 cites W3034530552 @default.
- W3136181114 cites W3034604951 @default.
- W3136181114 cites W3101765447 @default.
- W3136181114 cites W3107156787 @default.
- W3136181114 cites W3109908659 @default.
- W3136181114 cites W3110153602 @default.
- W3136181114 cites W612478963 @default.
- W3136181114 hasPublicationYear "2021" @default.
- W3136181114 type Work @default.
- W3136181114 sameAs 3136181114 @default.
- W3136181114 citedByCount "1" @default.
- W3136181114 countsByYear W31361811142021 @default.
- W3136181114 crossrefType "posted-content" @default.
- W3136181114 hasAuthorship W3136181114A5013549550 @default.
- W3136181114 hasAuthorship W3136181114A5029170925 @default.
- W3136181114 hasAuthorship W3136181114A5052772478 @default.
- W3136181114 hasAuthorship W3136181114A5060851814 @default.
- W3136181114 hasAuthorship W3136181114A5069033376 @default.
- W3136181114 hasAuthorship W3136181114A5084953118 @default.
- W3136181114 hasConcept C104114177 @default.
- W3136181114 hasConcept C108583219 @default.
- W3136181114 hasConcept C111919701 @default.
- W3136181114 hasConcept C11413529 @default.
- W3136181114 hasConcept C119857082 @default.
- W3136181114 hasConcept C136389625 @default.
- W3136181114 hasConcept C138673069 @default.
- W3136181114 hasConcept C138885662 @default.
- W3136181114 hasConcept C147168706 @default.
- W3136181114 hasConcept C153180895 @default.
- W3136181114 hasConcept C154945302 @default.
- W3136181114 hasConcept C162324750 @default.
- W3136181114 hasConcept C165696696 @default.
- W3136181114 hasConcept C176217482 @default.
- W3136181114 hasConcept C177264268 @default.
- W3136181114 hasConcept C199360897 @default.
- W3136181114 hasConcept C21547014 @default.
- W3136181114 hasConcept C2776401178 @default.
- W3136181114 hasConcept C2776760102 @default.
- W3136181114 hasConcept C38652104 @default.
- W3136181114 hasConcept C41008148 @default.
- W3136181114 hasConcept C41895202 @default.
- W3136181114 hasConcept C45374587 @default.
- W3136181114 hasConcept C50644808 @default.
- W3136181114 hasConceptScore W3136181114C104114177 @default.
- W3136181114 hasConceptScore W3136181114C108583219 @default.
- W3136181114 hasConceptScore W3136181114C111919701 @default.
- W3136181114 hasConceptScore W3136181114C11413529 @default.
- W3136181114 hasConceptScore W3136181114C119857082 @default.
- W3136181114 hasConceptScore W3136181114C136389625 @default.
- W3136181114 hasConceptScore W3136181114C138673069 @default.
- W3136181114 hasConceptScore W3136181114C138885662 @default.
- W3136181114 hasConceptScore W3136181114C147168706 @default.