Matches in SemOpenAlex for { <https://semopenalex.org/work/W3045257289> ?p ?o ?g. }
- W3045257289 abstract "We present a learning framework that learns to recover the 3D shape, pose and texture from a single image, trained on an image collection without any ground truth 3D shape, multi-view, camera viewpoints or keypoint supervision. We approach this highly under-constrained problem in a analysis by synthesis framework where the goal is to predict the likely shape, texture and camera viewpoint that could produce the image with various learned category-specific priors. Our particular contribution in this paper is a representation of the distribution over cameras, which we call camera-multiplex. Instead of picking a point estimate, we maintain a set of camera hypotheses that are optimized during training to best explain the image given the current shape and texture. We call our approach Unsupervised Category-Specific Mesh Reconstruction (U-CMR), and present qualitative and quantitative results on CUB, Pascal 3D and new web-scraped datasets. We obtain state-of-the-art camera prediction results and show that we can learn to predict diverse shapes and textures across objects using an image collection without any keypoint annotations or 3D ground truth. Project page: this https URL" @default.
- W3045257289 created "2020-07-29" @default.
- W3045257289 creator A5001594573 @default.
- W3045257289 creator A5026034978 @default.
- W3045257289 creator A5031491881 @default.
- W3045257289 date "2020-07-21" @default.
- W3045257289 modified "2023-09-27" @default.
- W3045257289 title "Shape and Viewpoint without Keypoints" @default.
- W3045257289 cites W1490482062 @default.
- W3045257289 cites W1797268635 @default.
- W3045257289 cites W183071939 @default.
- W3045257289 cites W1893912098 @default.
- W3045257289 cites W1920022804 @default.
- W3045257289 cites W1974956622 @default.
- W3045257289 cites W1977792424 @default.
- W3045257289 cites W1991264156 @default.
- W3045257289 cites W2066090933 @default.
- W3045257289 cites W2121299550 @default.
- W3045257289 cites W2237250383 @default.
- W3045257289 cites W2342277278 @default.
- W3045257289 cites W2395937778 @default.
- W3045257289 cites W2551540143 @default.
- W3045257289 cites W2557465155 @default.
- W3045257289 cites W2560722161 @default.
- W3045257289 cites W2576289912 @default.
- W3045257289 cites W2582734987 @default.
- W3045257289 cites W2772518193 @default.
- W3045257289 cites W2807725536 @default.
- W3045257289 cites W2890967717 @default.
- W3045257289 cites W2962729993 @default.
- W3045257289 cites W2962912205 @default.
- W3045257289 cites W2963022858 @default.
- W3045257289 cites W2963026643 @default.
- W3045257289 cites W2963238691 @default.
- W3045257289 cites W2963527086 @default.
- W3045257289 cites W2963627347 @default.
- W3045257289 cites W2963739349 @default.
- W3045257289 cites W2963823554 @default.
- W3045257289 cites W2963850211 @default.
- W3045257289 cites W2963893349 @default.
- W3045257289 cites W2963966978 @default.
- W3045257289 cites W2964137676 @default.
- W3045257289 cites W2970971581 @default.
- W3045257289 cites W2982024234 @default.
- W3045257289 cites W2990173985 @default.
- W3045257289 cites W2990578762 @default.
- W3045257289 cites W2991322756 @default.
- W3045257289 cites W3004414671 @default.
- W3045257289 cites W3009693647 @default.
- W3045257289 hasPublicationYear "2020" @default.
- W3045257289 type Work @default.
- W3045257289 sameAs 3045257289 @default.
- W3045257289 citedByCount "4" @default.
- W3045257289 countsByYear W30452572892020 @default.
- W3045257289 countsByYear W30452572892021 @default.
- W3045257289 crossrefType "posted-content" @default.
- W3045257289 hasAuthorship W3045257289A5001594573 @default.
- W3045257289 hasAuthorship W3045257289A5026034978 @default.
- W3045257289 hasAuthorship W3045257289A5031491881 @default.
- W3045257289 hasConcept C107673813 @default.
- W3045257289 hasConcept C115961682 @default.
- W3045257289 hasConcept C142362112 @default.
- W3045257289 hasConcept C146849305 @default.
- W3045257289 hasConcept C153180895 @default.
- W3045257289 hasConcept C153349607 @default.
- W3045257289 hasConcept C154945302 @default.
- W3045257289 hasConcept C177264268 @default.
- W3045257289 hasConcept C17744445 @default.
- W3045257289 hasConcept C177769412 @default.
- W3045257289 hasConcept C199360897 @default.
- W3045257289 hasConcept C199539241 @default.
- W3045257289 hasConcept C2524010 @default.
- W3045257289 hasConcept C2776035091 @default.
- W3045257289 hasConcept C2776359362 @default.
- W3045257289 hasConcept C28719098 @default.
- W3045257289 hasConcept C31972630 @default.
- W3045257289 hasConcept C33923547 @default.
- W3045257289 hasConcept C41008148 @default.
- W3045257289 hasConcept C75608658 @default.
- W3045257289 hasConcept C94625758 @default.
- W3045257289 hasConceptScore W3045257289C107673813 @default.
- W3045257289 hasConceptScore W3045257289C115961682 @default.
- W3045257289 hasConceptScore W3045257289C142362112 @default.
- W3045257289 hasConceptScore W3045257289C146849305 @default.
- W3045257289 hasConceptScore W3045257289C153180895 @default.
- W3045257289 hasConceptScore W3045257289C153349607 @default.
- W3045257289 hasConceptScore W3045257289C154945302 @default.
- W3045257289 hasConceptScore W3045257289C177264268 @default.
- W3045257289 hasConceptScore W3045257289C17744445 @default.
- W3045257289 hasConceptScore W3045257289C177769412 @default.
- W3045257289 hasConceptScore W3045257289C199360897 @default.
- W3045257289 hasConceptScore W3045257289C199539241 @default.
- W3045257289 hasConceptScore W3045257289C2524010 @default.
- W3045257289 hasConceptScore W3045257289C2776035091 @default.
- W3045257289 hasConceptScore W3045257289C2776359362 @default.
- W3045257289 hasConceptScore W3045257289C28719098 @default.
- W3045257289 hasConceptScore W3045257289C31972630 @default.
- W3045257289 hasConceptScore W3045257289C33923547 @default.
- W3045257289 hasConceptScore W3045257289C41008148 @default.
- W3045257289 hasConceptScore W3045257289C75608658 @default.