Matches in SemOpenAlex for { <https://semopenalex.org/work/W4387682004> ?p ?o ?g. }
Showing items 1 to 62 of
62
with 100 items per page.
- W4387682004 endingPage "12" @default.
- W4387682004 startingPage "1" @default.
- W4387682004 abstract "This paper presents a method to reconstruct high-quality textured 3D models from single images. Current methods rely on datasets with expensive annotations; multi-view images and their camera parameters. Our method relies on GAN generated multi-view image datasets which have a negligible annotation cost. However, they are not strictly multi-view consistent and sometimes GANs output distorted images. This results in degraded reconstruction qualities. In this work, to overcome these limitations of generated datasets, we have two main contributions which lead us to achieve state-of-the-art results on challenging objects: 1) A robust multi-stage learning scheme that gradually relies more on the models own predictions when calculating losses, 2) A novel adversarial learning pipeline with online pseudo-ground truth generations to achieve fine details. Our work provides a bridge from 2D supervisions of GAN models to 3D reconstruction models and removes the expensive annotation efforts. We show significant improvements over previous methods whether they were trained on GAN generated multi-view images or on real images with expensive annotations. Please visit our web-page for 3D visuals: https://research.nvidia.com/labs/adlr/progressive-3d-learning." @default.
- W4387682004 created "2023-10-17" @default.
- W4387682004 creator A5025713692 @default.
- W4387682004 creator A5041525280 @default.
- W4387682004 creator A5066242985 @default.
- W4387682004 creator A5086976460 @default.
- W4387682004 date "2023-01-01" @default.
- W4387682004 modified "2023-10-18" @default.
- W4387682004 title "Progressive Learning of 3D Reconstruction Network From 2D GAN Data" @default.
- W4387682004 doi "https://doi.org/10.1109/tpami.2023.3324806" @default.
- W4387682004 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/37844002" @default.
- W4387682004 hasPublicationYear "2023" @default.
- W4387682004 type Work @default.
- W4387682004 citedByCount "0" @default.
- W4387682004 crossrefType "journal-article" @default.
- W4387682004 hasAuthorship W4387682004A5025713692 @default.
- W4387682004 hasAuthorship W4387682004A5041525280 @default.
- W4387682004 hasAuthorship W4387682004A5066242985 @default.
- W4387682004 hasAuthorship W4387682004A5086976460 @default.
- W4387682004 hasConcept C108583219 @default.
- W4387682004 hasConcept C109950114 @default.
- W4387682004 hasConcept C119857082 @default.
- W4387682004 hasConcept C141379421 @default.
- W4387682004 hasConcept C146849305 @default.
- W4387682004 hasConcept C153180895 @default.
- W4387682004 hasConcept C154945302 @default.
- W4387682004 hasConcept C199360897 @default.
- W4387682004 hasConcept C2776321320 @default.
- W4387682004 hasConcept C31972630 @default.
- W4387682004 hasConcept C41008148 @default.
- W4387682004 hasConcept C43521106 @default.
- W4387682004 hasConceptScore W4387682004C108583219 @default.
- W4387682004 hasConceptScore W4387682004C109950114 @default.
- W4387682004 hasConceptScore W4387682004C119857082 @default.
- W4387682004 hasConceptScore W4387682004C141379421 @default.
- W4387682004 hasConceptScore W4387682004C146849305 @default.
- W4387682004 hasConceptScore W4387682004C153180895 @default.
- W4387682004 hasConceptScore W4387682004C154945302 @default.
- W4387682004 hasConceptScore W4387682004C199360897 @default.
- W4387682004 hasConceptScore W4387682004C2776321320 @default.
- W4387682004 hasConceptScore W4387682004C31972630 @default.
- W4387682004 hasConceptScore W4387682004C41008148 @default.
- W4387682004 hasConceptScore W4387682004C43521106 @default.
- W4387682004 hasLocation W43876820041 @default.
- W4387682004 hasLocation W43876820042 @default.
- W4387682004 hasOpenAccess W4387682004 @default.
- W4387682004 hasPrimaryLocation W43876820041 @default.
- W4387682004 hasRelatedWork W2144657392 @default.
- W4387682004 hasRelatedWork W2218034408 @default.
- W4387682004 hasRelatedWork W2263699433 @default.
- W4387682004 hasRelatedWork W2358755282 @default.
- W4387682004 hasRelatedWork W2361861616 @default.
- W4387682004 hasRelatedWork W2377979023 @default.
- W4387682004 hasRelatedWork W2392921965 @default.
- W4387682004 hasRelatedWork W2625833328 @default.
- W4387682004 hasRelatedWork W4323323175 @default.
- W4387682004 hasRelatedWork W4386065843 @default.
- W4387682004 isParatext "false" @default.
- W4387682004 isRetracted "false" @default.
- W4387682004 workType "article" @default.