Matches in SemOpenAlex for { <https://semopenalex.org/work/W3194356231> ?p ?o ?g. }
- W3194356231 abstract "Recent work has made significant progress in learning object meshes with weak supervision. Soft Rasterization methods have achieved accurate 3D reconstruction from 2D images with viewpoint supervision only. In this work, we further reduce the labeling effort by allowing such 3D reconstruction methods leverage unlabeled images. In order to obtain the viewpoints for these unlabeled images, we propose to use a Siamese network that takes two images as input and outputs whether they correspond to the same viewpoint. During training, we minimize the cross entropy loss to maximize the probability of predicting whether a pair of images belong to the same viewpoint or not. To get the viewpoint of a new image, we compare it against different viewpoints obtained from the training samples and select the viewpoint with the highest matching probability. We finally label the unlabeled images with the most confident predicted viewpoint and train a deep network that has a differentiable rasterization layer. Our experiments show that even labeling only two objects yields significant improvement in IoU for ShapeNet when leveraging unlabeled examples. Code is available at this https URL." @default.
- W3194356231 created "2021-08-30" @default.
- W3194356231 creator A5004903288 @default.
- W3194356231 creator A5028905152 @default.
- W3194356231 creator A5053824527 @default.
- W3194356231 creator A5077092609 @default.
- W3194356231 date "2021-08-21" @default.
- W3194356231 modified "2023-09-27" @default.
- W3194356231 title "SSR: Semi-supervised Soft Rasterizer for single-view 2D to 3D Reconstruction" @default.
- W3194356231 cites W1677182931 @default.
- W3194356231 cites W1983320747 @default.
- W3194356231 cites W2066090933 @default.
- W3194356231 cites W2157364932 @default.
- W3194356231 cites W2174133987 @default.
- W3194356231 cites W2190691619 @default.
- W3194356231 cites W2342277278 @default.
- W3194356231 cites W2469266052 @default.
- W3194356231 cites W2551540143 @default.
- W3194356231 cites W2555618208 @default.
- W3194356231 cites W2738835886 @default.
- W3194356231 cites W2739245956 @default.
- W3194356231 cites W2886253469 @default.
- W3194356231 cites W2887338480 @default.
- W3194356231 cites W2902690438 @default.
- W3194356231 cites W2929298487 @default.
- W3194356231 cites W2948124424 @default.
- W3194356231 cites W2962778872 @default.
- W3194356231 cites W2962912205 @default.
- W3194356231 cites W2963453931 @default.
- W3194356231 cites W2963527086 @default.
- W3194356231 cites W2963926543 @default.
- W3194356231 cites W2964137676 @default.
- W3194356231 cites W2964472976 @default.
- W3194356231 cites W2988472682 @default.
- W3194356231 cites W2990173985 @default.
- W3194356231 cites W2990578762 @default.
- W3194356231 cites W3009561768 @default.
- W3194356231 cites W3034499993 @default.
- W3194356231 cites W3034978746 @default.
- W3194356231 cites W3035060554 @default.
- W3194356231 cites W3035160371 @default.
- W3194356231 cites W3039448353 @default.
- W3194356231 cites W3091905774 @default.
- W3194356231 cites W3096831136 @default.
- W3194356231 cites W3098589080 @default.
- W3194356231 cites W3099570996 @default.
- W3194356231 cites W3107127767 @default.
- W3194356231 cites W3110214837 @default.
- W3194356231 cites W3144483443 @default.
- W3194356231 hasPublicationYear "2021" @default.
- W3194356231 type Work @default.
- W3194356231 sameAs 3194356231 @default.
- W3194356231 citedByCount "0" @default.
- W3194356231 crossrefType "posted-content" @default.
- W3194356231 hasAuthorship W3194356231A5004903288 @default.
- W3194356231 hasAuthorship W3194356231A5028905152 @default.
- W3194356231 hasAuthorship W3194356231A5053824527 @default.
- W3194356231 hasAuthorship W3194356231A5077092609 @default.
- W3194356231 hasConcept C105795698 @default.
- W3194356231 hasConcept C106301342 @default.
- W3194356231 hasConcept C111919701 @default.
- W3194356231 hasConcept C115961682 @default.
- W3194356231 hasConcept C121332964 @default.
- W3194356231 hasConcept C121684516 @default.
- W3194356231 hasConcept C134306372 @default.
- W3194356231 hasConcept C142362112 @default.
- W3194356231 hasConcept C153083717 @default.
- W3194356231 hasConcept C153180895 @default.
- W3194356231 hasConcept C153349607 @default.
- W3194356231 hasConcept C154945302 @default.
- W3194356231 hasConcept C165064840 @default.
- W3194356231 hasConcept C177264268 @default.
- W3194356231 hasConcept C199360897 @default.
- W3194356231 hasConcept C202615002 @default.
- W3194356231 hasConcept C2776035091 @default.
- W3194356231 hasConcept C2776760102 @default.
- W3194356231 hasConcept C31487907 @default.
- W3194356231 hasConcept C31972630 @default.
- W3194356231 hasConcept C33923547 @default.
- W3194356231 hasConcept C41008148 @default.
- W3194356231 hasConcept C43126263 @default.
- W3194356231 hasConcept C62520636 @default.
- W3194356231 hasConceptScore W3194356231C105795698 @default.
- W3194356231 hasConceptScore W3194356231C106301342 @default.
- W3194356231 hasConceptScore W3194356231C111919701 @default.
- W3194356231 hasConceptScore W3194356231C115961682 @default.
- W3194356231 hasConceptScore W3194356231C121332964 @default.
- W3194356231 hasConceptScore W3194356231C121684516 @default.
- W3194356231 hasConceptScore W3194356231C134306372 @default.
- W3194356231 hasConceptScore W3194356231C142362112 @default.
- W3194356231 hasConceptScore W3194356231C153083717 @default.
- W3194356231 hasConceptScore W3194356231C153180895 @default.
- W3194356231 hasConceptScore W3194356231C153349607 @default.
- W3194356231 hasConceptScore W3194356231C154945302 @default.
- W3194356231 hasConceptScore W3194356231C165064840 @default.
- W3194356231 hasConceptScore W3194356231C177264268 @default.
- W3194356231 hasConceptScore W3194356231C199360897 @default.
- W3194356231 hasConceptScore W3194356231C202615002 @default.
- W3194356231 hasConceptScore W3194356231C2776035091 @default.
- W3194356231 hasConceptScore W3194356231C2776760102 @default.