Matches in SemOpenAlex for { <https://semopenalex.org/work/W4312795764> ?p ?o ?g. }
- W4312795764 abstract "Current deep neural network approaches for camera pose estimation rely on scene structure for 3D motion estimation, but this decreases the robustness and thereby makes cross-dataset generalization difficult. In contrast, classical approaches to structure from motion estimate 3D motion utilizing optical flow and then compute depth. Their accuracy, however, depends strongly on the quality of the optical flow. To avoid this issue, direct methods have been proposed, which separate 3D motion from depth estimation, but compute 3D motion using only image gradients in the form of normal flow. In this paper, we introduce a network NFlowNet, for normal flow estimation which is used to enforce robust and direct constraints. In particular, normal flow is used to estimate relative camera pose based on the cheirality (depth positivity) constraint. We achieve this by formulating the optimization problem as a differentiable cheirality layer, which allows for end-to-end learning of camera pose. We perform extensive qualitative and quantitative evaluation of the proposed DiffPoseNet's sensitivity to noise and its generalization across datasets. We compare our approach to existing state-of-the-art methods on KITTI, TartanAir, and TUM-RGBD datasets." @default.
- W4312795764 created "2023-01-05" @default.
- W4312795764 creator A5004489033 @default.
- W4312795764 creator A5036912867 @default.
- W4312795764 creator A5055752014 @default.
- W4312795764 creator A5075055783 @default.
- W4312795764 creator A5083553427 @default.
- W4312795764 date "2022-06-01" @default.
- W4312795764 modified "2023-10-01" @default.
- W4312795764 title "DiffPoseNet: Direct Differentiable Camera Pose Estimation" @default.
- W4312795764 cites W1572666543 @default.
- W4312795764 cites W1602041476 @default.
- W4312795764 cites W1988849934 @default.
- W4312795764 cites W2000359198 @default.
- W4312795764 cites W2012929831 @default.
- W4312795764 cites W2015996585 @default.
- W4312795764 cites W2020704715 @default.
- W4312795764 cites W2021851106 @default.
- W4312795764 cites W2053088632 @default.
- W4312795764 cites W2072587497 @default.
- W4312795764 cites W2085926948 @default.
- W4312795764 cites W2086532466 @default.
- W4312795764 cites W2100315781 @default.
- W4312795764 cites W2127490113 @default.
- W4312795764 cites W2131747574 @default.
- W4312795764 cites W2139782508 @default.
- W4312795764 cites W2150066425 @default.
- W4312795764 cites W2152671441 @default.
- W4312795764 cites W2157301061 @default.
- W4312795764 cites W2163082602 @default.
- W4312795764 cites W2179278902 @default.
- W4312795764 cites W2200124539 @default.
- W4312795764 cites W2292391751 @default.
- W4312795764 cites W2548527721 @default.
- W4312795764 cites W2560474170 @default.
- W4312795764 cites W2598706937 @default.
- W4312795764 cites W2605111497 @default.
- W4312795764 cites W2963583471 @default.
- W4312795764 cites W2963782415 @default.
- W4312795764 cites W2963891416 @default.
- W4312795764 cites W2964314455 @default.
- W4312795764 cites W2983371255 @default.
- W4312795764 cites W3003957707 @default.
- W4312795764 cites W3034364596 @default.
- W4312795764 cites W3100280854 @default.
- W4312795764 cites W3101197114 @default.
- W4312795764 cites W3102327032 @default.
- W4312795764 cites W3103648783 @default.
- W4312795764 cites W3107714803 @default.
- W4312795764 cites W3126955704 @default.
- W4312795764 cites W3129709393 @default.
- W4312795764 cites W3131747507 @default.
- W4312795764 cites W3132270109 @default.
- W4312795764 cites W3175647732 @default.
- W4312795764 cites W764651262 @default.
- W4312795764 doi "https://doi.org/10.1109/cvpr52688.2022.00672" @default.
- W4312795764 hasPublicationYear "2022" @default.
- W4312795764 type Work @default.
- W4312795764 citedByCount "5" @default.
- W4312795764 countsByYear W43127957642022 @default.
- W4312795764 countsByYear W43127957642023 @default.
- W4312795764 crossrefType "proceedings-article" @default.
- W4312795764 hasAuthorship W4312795764A5004489033 @default.
- W4312795764 hasAuthorship W4312795764A5036912867 @default.
- W4312795764 hasAuthorship W4312795764A5055752014 @default.
- W4312795764 hasAuthorship W4312795764A5075055783 @default.
- W4312795764 hasAuthorship W4312795764A5083553427 @default.
- W4312795764 hasBestOaLocation W43127957642 @default.
- W4312795764 hasConcept C10161872 @default.
- W4312795764 hasConcept C104317684 @default.
- W4312795764 hasConcept C115961682 @default.
- W4312795764 hasConcept C134306372 @default.
- W4312795764 hasConcept C146159030 @default.
- W4312795764 hasConcept C153180895 @default.
- W4312795764 hasConcept C154945302 @default.
- W4312795764 hasConcept C155542232 @default.
- W4312795764 hasConcept C177148314 @default.
- W4312795764 hasConcept C185592680 @default.
- W4312795764 hasConcept C202615002 @default.
- W4312795764 hasConcept C31972630 @default.
- W4312795764 hasConcept C33923547 @default.
- W4312795764 hasConcept C41008148 @default.
- W4312795764 hasConcept C50644808 @default.
- W4312795764 hasConcept C52102323 @default.
- W4312795764 hasConcept C55493867 @default.
- W4312795764 hasConcept C63479239 @default.
- W4312795764 hasConceptScore W4312795764C10161872 @default.
- W4312795764 hasConceptScore W4312795764C104317684 @default.
- W4312795764 hasConceptScore W4312795764C115961682 @default.
- W4312795764 hasConceptScore W4312795764C134306372 @default.
- W4312795764 hasConceptScore W4312795764C146159030 @default.
- W4312795764 hasConceptScore W4312795764C153180895 @default.
- W4312795764 hasConceptScore W4312795764C154945302 @default.
- W4312795764 hasConceptScore W4312795764C155542232 @default.
- W4312795764 hasConceptScore W4312795764C177148314 @default.
- W4312795764 hasConceptScore W4312795764C185592680 @default.
- W4312795764 hasConceptScore W4312795764C202615002 @default.
- W4312795764 hasConceptScore W4312795764C31972630 @default.
- W4312795764 hasConceptScore W4312795764C33923547 @default.
- W4312795764 hasConceptScore W4312795764C41008148 @default.