FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3133571665> ?p ?o ?g. }

• W3133571665 abstract "We present self-supervised geometric perception (SGP), the first general framework to learn a feature descriptor for correspondence matching without any ground-truth geometric model labels (e.g., camera poses, rigid transformations). Our first contribution is to formulate geometric perception as an optimization problem that jointly optimizes the feature descriptor and the geometric models given a large corpus of visual measurements (e.g., images, point clouds). Under this optimization formulation, we show that two important streams of research in vision, namely robust model fitting and deep feature learning, correspond to optimizing one block of the unknown variables while fixing the other block. This analysis naturally leads to our second contribution -- the SGP algorithm that performs alternating minimization to solve the joint optimization. SGP iteratively executes two meta-algorithms: a teacher that performs robust model fitting given learned features to generate geometric pseudo-labels, and a student that performs deep feature learning under noisy supervision of the pseudo-labels. As a third contribution, we apply SGP to two perception problems on large-scale real datasets, namely relative camera pose estimation on MegaDepth and point cloud registration on 3DMatch. We demonstrate that SGP achieves state-of-the-art performance that is on-par or superior to the supervised oracles trained using ground-truth labels." @default.
• W3133571665 created "2021-03-15" @default.
• W3133571665 creator A5006181255 @default.
• W3133571665 creator A5042157108 @default.
• W3133571665 creator A5080636854 @default.
• W3133571665 creator A5086692495 @default.
• W3133571665 date "2021-03-04" @default.
• W3133571665 modified "2023-10-16" @default.
• W3133571665 title "Self-supervised Geometric Perception" @default.
• W3133571665 cites W124242857 @default.
• W3133571665 cites W132147841 @default.
• W3133571665 cites W1907417778 @default.
• W3133571665 cites W1957167950 @default.
• W3133571665 cites W1989625560 @default.
• W3133571665 cites W1993267444 @default.
• W3133571665 cites W2019417606 @default.
• W3133571665 cites W2021851106 @default.
• W3133571665 cites W2033819227 @default.
• W3133571665 cites W2049981393 @default.
• W3133571665 cites W2085261163 @default.
• W3133571665 cites W2097649661 @default.
• W3133571665 cites W2109635530 @default.
• W3133571665 cites W2133442958 @default.
• W3133571665 cites W2139103507 @default.
• W3133571665 cites W2144824356 @default.
• W3133571665 cites W2145494108 @default.
• W3133571665 cites W2151290401 @default.
• W3133571665 cites W2159719921 @default.
• W3133571665 cites W2160821342 @default.
• W3133571665 cites W2181274189 @default.
• W3133571665 cites W2205514457 @default.
• W3133571665 cites W2253156915 @default.
• W3133571665 cites W2326925005 @default.
• W3133571665 cites W2335698258 @default.
• W3133571665 cites W2519911873 @default.
• W3133571665 cites W2558625610 @default.
• W3133571665 cites W2562843573 @default.
• W3133571665 cites W2592936284 @default.
• W3133571665 cites W2593721734 @default.
• W3133571665 cites W2594519801 @default.
• W3133571665 cites W2609883120 @default.
• W3133571665 cites W2737260104 @default.
• W3133571665 cites W2768879211 @default.
• W3133571665 cites W2769591697 @default.
• W3133571665 cites W2776330782 @default.
• W3133571665 cites W2786036844 @default.
• W3133571665 cites W2795912842 @default.
• W3133571665 cites W2798766386 @default.
• W3133571665 cites W2883357174 @default.
• W3133571665 cites W2890097289 @default.
• W3133571665 cites W2891820683 @default.
• W3133571665 cites W2962729173 @default.
• W3133571665 cites W2962828767 @default.
• W3133571665 cites W2962867954 @default.
• W3133571665 cites W2962958090 @default.
• W3133571665 cites W2963188159 @default.
• W3133571665 cites W2963654727 @default.
• W3133571665 cites W2963666542 @default.
• W3133571665 cites W2963756608 @default.
• W3133571665 cites W2963760790 @default.
• W3133571665 cites W2964000117 @default.
• W3133571665 cites W2964014140 @default.
• W3133571665 cites W2964047820 @default.
• W3133571665 cites W2967043539 @default.
• W3133571665 cites W2971272421 @default.
• W3133571665 cites W2973258311 @default.
• W3133571665 cites W2981378444 @default.
• W3133571665 cites W2981556384 @default.
• W3133571665 cites W2981874330 @default.
• W3133571665 cites W2981995220 @default.
• W3133571665 cites W2985775862 @default.
• W3133571665 cites W2986382673 @default.
• W3133571665 cites W2988085408 @default.
• W3133571665 cites W2990111328 @default.
• W3133571665 cites W2990613095 @default.
• W3133571665 cites W3003372926 @default.
• W3133571665 cites W3023371261 @default.
• W3133571665 cites W3034373437 @default.
• W3133571665 cites W3034494757 @default.
• W3133571665 cites W3034675048 @default.
• W3133571665 cites W3035030518 @default.
• W3133571665 cites W3035056458 @default.
• W3133571665 cites W3035160371 @default.
• W3133571665 cites W3035272603 @default.
• W3133571665 cites W3043075211 @default.
• W3133571665 cites W3045125647 @default.
• W3133571665 cites W3045905189 @default.
• W3133571665 cites W3067981314 @default.
• W3133571665 cites W3093386634 @default.
• W3133571665 cites W3102327032 @default.
• W3133571665 cites W3105459032 @default.
• W3133571665 cites W3107540572 @default.
• W3133571665 cites W3111598592 @default.
• W3133571665 cites W3120044914 @default.
• W3133571665 cites W3124420883 @default.
• W3133571665 cites W3131747507 @default.
• W3133571665 cites W3207682209 @default.
• W3133571665 doi "https://doi.org/10.48550/arxiv.2103.03114" @default.
• W3133571665 hasPublicationYear "2021" @default.
• W3133571665 type Work @default.

• next