Matches in SemOpenAlex for { <https://semopenalex.org/work/W3203760490> ?p ?o ?g. }
Showing items 1 to 86 of
86
with 100 items per page.
- W3203760490 abstract "Deep CNN-based methods have so far achieved the state of the art results in multi-view 3D object reconstruction. Despite the considerable progress, the two core modules of these methods - multi-view feature extraction and fusion, are usually investigated separately, and the object relations in different views are rarely explored. In this paper, inspired by the recent great success in self-attention-based Transformer models, we reformulate the multi-view 3D reconstruction as a sequence-to-sequence prediction problem and propose a new framework named 3D Volume Transformer (VolT) for such a task. Unlike previous CNN-based methods using a separate design, we unify the feature extraction and view fusion in a single Transformer network. A natural advantage of our design lies in the exploration of view-to-view relationships using self-attention among multiple unordered inputs. On ShapeNet - a large-scale 3D reconstruction benchmark dataset, our method achieves a new state-of-the-art accuracy in multi-view reconstruction with fewer parameters ($70%$ less) than other CNN-based methods. Experimental results also suggest the strong scaling capability of our method. Our code will be made publicly available." @default.
- W3203760490 created "2021-10-11" @default.
- W3203760490 creator A5028143781 @default.
- W3203760490 creator A5028375560 @default.
- W3203760490 creator A5036468132 @default.
- W3203760490 creator A5042295331 @default.
- W3203760490 creator A5057637393 @default.
- W3203760490 creator A5058380236 @default.
- W3203760490 creator A5083857025 @default.
- W3203760490 creator A5088611151 @default.
- W3203760490 date "2021-03-23" @default.
- W3203760490 modified "2023-09-25" @default.
- W3203760490 title "Multi-view 3D Reconstruction with Transformer" @default.
- W3203760490 cites W1686810756 @default.
- W3203760490 cites W2194775991 @default.
- W3203760490 cites W2342277278 @default.
- W3203760490 cites W2798856139 @default.
- W3203760490 cites W2908510526 @default.
- W3203760490 cites W2962778872 @default.
- W3203760490 cites W2963221299 @default.
- W3203760490 cites W2963341956 @default.
- W3203760490 cites W2963403868 @default.
- W3203760490 cites W2963966978 @default.
- W3203760490 cites W2968257580 @default.
- W3203760490 cites W2971278627 @default.
- W3203760490 cites W3099201369 @default.
- W3203760490 cites W3105863736 @default.
- W3203760490 cites W3173531806 @default.
- W3203760490 cites W3186630079 @default.
- W3203760490 doi "https://doi.org/10.48550/arxiv.2103.12957" @default.
- W3203760490 hasPublicationYear "2021" @default.
- W3203760490 type Work @default.
- W3203760490 sameAs 3203760490 @default.
- W3203760490 citedByCount "1" @default.
- W3203760490 countsByYear W32037604902021 @default.
- W3203760490 crossrefType "posted-content" @default.
- W3203760490 hasAuthorship W3203760490A5028143781 @default.
- W3203760490 hasAuthorship W3203760490A5028375560 @default.
- W3203760490 hasAuthorship W3203760490A5036468132 @default.
- W3203760490 hasAuthorship W3203760490A5042295331 @default.
- W3203760490 hasAuthorship W3203760490A5057637393 @default.
- W3203760490 hasAuthorship W3203760490A5058380236 @default.
- W3203760490 hasAuthorship W3203760490A5083857025 @default.
- W3203760490 hasAuthorship W3203760490A5088611151 @default.
- W3203760490 hasBestOaLocation W32037604901 @default.
- W3203760490 hasConcept C111919701 @default.
- W3203760490 hasConcept C119599485 @default.
- W3203760490 hasConcept C127413603 @default.
- W3203760490 hasConcept C13280743 @default.
- W3203760490 hasConcept C153180895 @default.
- W3203760490 hasConcept C154945302 @default.
- W3203760490 hasConcept C165801399 @default.
- W3203760490 hasConcept C177774035 @default.
- W3203760490 hasConcept C185798385 @default.
- W3203760490 hasConcept C205649164 @default.
- W3203760490 hasConcept C41008148 @default.
- W3203760490 hasConcept C66322947 @default.
- W3203760490 hasConceptScore W3203760490C111919701 @default.
- W3203760490 hasConceptScore W3203760490C119599485 @default.
- W3203760490 hasConceptScore W3203760490C127413603 @default.
- W3203760490 hasConceptScore W3203760490C13280743 @default.
- W3203760490 hasConceptScore W3203760490C153180895 @default.
- W3203760490 hasConceptScore W3203760490C154945302 @default.
- W3203760490 hasConceptScore W3203760490C165801399 @default.
- W3203760490 hasConceptScore W3203760490C177774035 @default.
- W3203760490 hasConceptScore W3203760490C185798385 @default.
- W3203760490 hasConceptScore W3203760490C205649164 @default.
- W3203760490 hasConceptScore W3203760490C41008148 @default.
- W3203760490 hasConceptScore W3203760490C66322947 @default.
- W3203760490 hasLocation W32037604901 @default.
- W3203760490 hasOpenAccess W3203760490 @default.
- W3203760490 hasPrimaryLocation W32037604901 @default.
- W3203760490 hasRelatedWork W112744582 @default.
- W3203760490 hasRelatedWork W1485630101 @default.
- W3203760490 hasRelatedWork W1594844924 @default.
- W3203760490 hasRelatedWork W1649510843 @default.
- W3203760490 hasRelatedWork W2143670980 @default.
- W3203760490 hasRelatedWork W2315101603 @default.
- W3203760490 hasRelatedWork W2326649917 @default.
- W3203760490 hasRelatedWork W2498017833 @default.
- W3203760490 hasRelatedWork W2909382770 @default.
- W3203760490 hasRelatedWork W4286750849 @default.
- W3203760490 isParatext "false" @default.
- W3203760490 isRetracted "false" @default.
- W3203760490 magId "3203760490" @default.
- W3203760490 workType "article" @default.