Matches in SemOpenAlex for { <https://semopenalex.org/work/W4381930141> ?p ?o ?g. }
Showing items 1 to 94 of
94
with 100 items per page.
- W4381930141 endingPage "2769" @default.
- W4381930141 startingPage "2757" @default.
- W4381930141 abstract "A popular approach to estimate scene flow is to utilize point cloud data from various Lidar scans. However, there is little attention to learning 3D motion from camera images. Learning scene flow from a monocular camera remains a challenging task due to its ill-posedness as well as lack of annotated data. Self-supervised methods demonstrate learning scene flow estimation from unlabeled data, yet their accuracy lags behind (semi-)supervised methods. In this paper, we introduce a self-supervised monocular scene flow method that substantially improves the accuracy over the previous approaches. Based on RAFT, a state-of-the-art optical flow model, we design a new decoder to iteratively update 3D motion fields and disparity maps simultaneously. Furthermore, we propose an enhanced upsampling layer and a disparity initialization technique, which overall further improves accuracy up to 7.2%. Our method achieves state-of-the-art accuracy among all self-supervised monocular scene flow methods, improving accuracy by 34.2%. Our fine-tuned model outperforms the best previous semi-supervised method with 228 times faster runtime. Code will be publicly available to ensure reproducibility." @default.
- W4381930141 created "2023-06-25" @default.
- W4381930141 creator A5000312477 @default.
- W4381930141 creator A5005363574 @default.
- W4381930141 creator A5030781794 @default.
- W4381930141 date "2023-06-24" @default.
- W4381930141 modified "2023-09-27" @default.
- W4381930141 title "RAFT-MSF: Self-Supervised Monocular Scene Flow Using Recurrent Optimizer" @default.
- W4381930141 cites W1508115313 @default.
- W4381930141 cites W1968545482 @default.
- W4381930141 cites W2129671742 @default.
- W4381930141 cites W2133665775 @default.
- W4381930141 cites W2520707372 @default.
- W4381930141 cites W2810554001 @default.
- W4381930141 cites W2890949887 @default.
- W4381930141 cites W2962771259 @default.
- W4381930141 cites W2963265330 @default.
- W4381930141 cites W2963583471 @default.
- W4381930141 cites W2963782415 @default.
- W4381930141 cites W2963891416 @default.
- W4381930141 cites W2964677288 @default.
- W4381930141 cites W2964968086 @default.
- W4381930141 cites W2971686478 @default.
- W4381930141 cites W2981518351 @default.
- W4381930141 cites W2981732213 @default.
- W4381930141 cites W2985178537 @default.
- W4381930141 cites W2985775862 @default.
- W4381930141 cites W3010090948 @default.
- W4381930141 cites W3034193257 @default.
- W4381930141 cites W3034642839 @default.
- W4381930141 cites W3097139398 @default.
- W4381930141 cites W3113357670 @default.
- W4381930141 cites W3173561529 @default.
- W4381930141 cites W3174581459 @default.
- W4381930141 cites W3186096360 @default.
- W4381930141 cites W3187838476 @default.
- W4381930141 cites W3203943632 @default.
- W4381930141 cites W3204559841 @default.
- W4381930141 doi "https://doi.org/10.1007/s11263-023-01828-4" @default.
- W4381930141 hasPublicationYear "2023" @default.
- W4381930141 type Work @default.
- W4381930141 citedByCount "0" @default.
- W4381930141 crossrefType "journal-article" @default.
- W4381930141 hasAuthorship W4381930141A5000312477 @default.
- W4381930141 hasAuthorship W4381930141A5005363574 @default.
- W4381930141 hasAuthorship W4381930141A5030781794 @default.
- W4381930141 hasBestOaLocation W43819301412 @default.
- W4381930141 hasConcept C110384440 @default.
- W4381930141 hasConcept C114466953 @default.
- W4381930141 hasConcept C115961682 @default.
- W4381930141 hasConcept C131979681 @default.
- W4381930141 hasConcept C136389625 @default.
- W4381930141 hasConcept C153180895 @default.
- W4381930141 hasConcept C154945302 @default.
- W4381930141 hasConcept C155542232 @default.
- W4381930141 hasConcept C199360897 @default.
- W4381930141 hasConcept C31972630 @default.
- W4381930141 hasConcept C41008148 @default.
- W4381930141 hasConcept C50644808 @default.
- W4381930141 hasConcept C65909025 @default.
- W4381930141 hasConceptScore W4381930141C110384440 @default.
- W4381930141 hasConceptScore W4381930141C114466953 @default.
- W4381930141 hasConceptScore W4381930141C115961682 @default.
- W4381930141 hasConceptScore W4381930141C131979681 @default.
- W4381930141 hasConceptScore W4381930141C136389625 @default.
- W4381930141 hasConceptScore W4381930141C153180895 @default.
- W4381930141 hasConceptScore W4381930141C154945302 @default.
- W4381930141 hasConceptScore W4381930141C155542232 @default.
- W4381930141 hasConceptScore W4381930141C199360897 @default.
- W4381930141 hasConceptScore W4381930141C31972630 @default.
- W4381930141 hasConceptScore W4381930141C41008148 @default.
- W4381930141 hasConceptScore W4381930141C50644808 @default.
- W4381930141 hasConceptScore W4381930141C65909025 @default.
- W4381930141 hasIssue "11" @default.
- W4381930141 hasLocation W43819301411 @default.
- W4381930141 hasLocation W43819301412 @default.
- W4381930141 hasOpenAccess W4381930141 @default.
- W4381930141 hasPrimaryLocation W43819301411 @default.
- W4381930141 hasRelatedWork W1970467378 @default.
- W4381930141 hasRelatedWork W2099272513 @default.
- W4381930141 hasRelatedWork W2156208915 @default.
- W4381930141 hasRelatedWork W2592071824 @default.
- W4381930141 hasRelatedWork W274991966 @default.
- W4381930141 hasRelatedWork W2992415220 @default.
- W4381930141 hasRelatedWork W3002248669 @default.
- W4381930141 hasRelatedWork W3198126099 @default.
- W4381930141 hasRelatedWork W4229080984 @default.
- W4381930141 hasRelatedWork W4281553171 @default.
- W4381930141 hasVolume "131" @default.
- W4381930141 isParatext "false" @default.
- W4381930141 isRetracted "false" @default.
- W4381930141 workType "article" @default.