Matches in SemOpenAlex for { <https://semopenalex.org/work/W3194086169> ?p ?o ?g. }
- W3194086169 abstract "We present an unsupervised learning framework for decomposing images into layers of automatically discovered object models. Contrary to recent approaches that model image layers with autoencoder networks, we represent them as explicit transformations of a small set of prototypical images. Our model has three main components: (i) a set of object prototypes in the form of learnable images with a transparency channel, which we refer to as sprites; (ii) differentiable parametric functions predicting occlusions and transformation parameters necessary to instantiate the sprites in a given image; (iii) a layered image formation model with occlusion for compositing these instances into complete images including background. By jointly learning the sprites and occlusion/transformation predictors to reconstruct images, our approach not only yields accurate layered image decompositions, but also identifies object categories and instance parameters. We first validate our approach by providing results on par with the state of the art on standard multi-object synthetic benchmarks (Tetrominoes, Multi-dSprites, CLEVR6). We then demonstrate the applicability of our model to real images in tasks that include clustering (SVHN, GTSRB), cosegmentation (Weizmann Horse) and object discovery from unfiltered social network images. To the best of our knowledge, our approach is the first layered image decomposition algorithm that learns an explicit and shared concept of object type, and is robust enough to be applied to real images." @default.
- W3194086169 created "2021-08-30" @default.
- W3194086169 creator A5011414147 @default.
- W3194086169 creator A5039530212 @default.
- W3194086169 creator A5061634216 @default.
- W3194086169 creator A5075666099 @default.
- W3194086169 date "2021-10-11" @default.
- W3194086169 modified "2023-10-14" @default.
- W3194086169 title "Unsupervised Layered Image Decomposition into Object Prototypes" @default.
- W3194086169 cites W1548595933 @default.
- W3194086169 cites W1829102423 @default.
- W3194086169 cites W1963920598 @default.
- W3194086169 cites W1996140089 @default.
- W3194086169 cites W2049633694 @default.
- W3194086169 cites W2067713319 @default.
- W3194086169 cites W2079550083 @default.
- W3194086169 cites W2081589848 @default.
- W3194086169 cites W2086052791 @default.
- W3194086169 cites W2099471712 @default.
- W3194086169 cites W2103658758 @default.
- W3194086169 cites W2112301665 @default.
- W3194086169 cites W2115665694 @default.
- W3194086169 cites W2119474464 @default.
- W3194086169 cites W2123782934 @default.
- W3194086169 cites W2127218421 @default.
- W3194086169 cites W2128764958 @default.
- W3194086169 cites W2131782326 @default.
- W3194086169 cites W2133774033 @default.
- W3194086169 cites W2134529554 @default.
- W3194086169 cites W2134665698 @default.
- W3194086169 cites W2135265773 @default.
- W3194086169 cites W2137821246 @default.
- W3194086169 cites W2153760331 @default.
- W3194086169 cites W2160847065 @default.
- W3194086169 cites W2185466002 @default.
- W3194086169 cites W2194775991 @default.
- W3194086169 cites W2222512263 @default.
- W3194086169 cites W2262458959 @default.
- W3194086169 cites W2269947760 @default.
- W3194086169 cites W2274801302 @default.
- W3194086169 cites W2295160225 @default.
- W3194086169 cites W2296591648 @default.
- W3194086169 cites W2335728318 @default.
- W3194086169 cites W2533113478 @default.
- W3194086169 cites W2547875792 @default.
- W3194086169 cites W2548228487 @default.
- W3194086169 cites W2561715562 @default.
- W3194086169 cites W2592101326 @default.
- W3194086169 cites W2593814746 @default.
- W3194086169 cites W2739010165 @default.
- W3194086169 cites W2769257693 @default.
- W3194086169 cites W2769443878 @default.
- W3194086169 cites W2788552663 @default.
- W3194086169 cites W2807242871 @default.
- W3194086169 cites W2903141607 @default.
- W3194086169 cites W2903585023 @default.
- W3194086169 cites W2911448865 @default.
- W3194086169 cites W2913954345 @default.
- W3194086169 cites W2946952476 @default.
- W3194086169 cites W2962889261 @default.
- W3194086169 cites W2963370555 @default.
- W3194086169 cites W2963403868 @default.
- W3194086169 cites W2963717490 @default.
- W3194086169 cites W2963951231 @default.
- W3194086169 cites W2964082390 @default.
- W3194086169 cites W2964121744 @default.
- W3194086169 cites W2964213104 @default.
- W3194086169 cites W2966661 @default.
- W3194086169 cites W2970653639 @default.
- W3194086169 cites W2971040589 @default.
- W3194086169 cites W2971044234 @default.
- W3194086169 cites W2979654309 @default.
- W3194086169 cites W2981585936 @default.
- W3194086169 cites W2990500698 @default.
- W3194086169 cites W2990844506 @default.
- W3194086169 cites W2994971263 @default.
- W3194086169 cites W3000266068 @default.
- W3194086169 cites W3015218297 @default.
- W3194086169 cites W3020457430 @default.
- W3194086169 cites W3098902916 @default.
- W3194086169 cites W3102696055 @default.
- W3194086169 cites W3106670560 @default.
- W3194086169 cites W3110446398 @default.
- W3194086169 cites W603908379 @default.
- W3194086169 hasPublicationYear "2021" @default.
- W3194086169 type Work @default.
- W3194086169 sameAs 3194086169 @default.
- W3194086169 citedByCount "0" @default.
- W3194086169 crossrefType "proceedings-article" @default.
- W3194086169 hasAuthorship W3194086169A5011414147 @default.
- W3194086169 hasAuthorship W3194086169A5039530212 @default.
- W3194086169 hasAuthorship W3194086169A5061634216 @default.
- W3194086169 hasAuthorship W3194086169A5075666099 @default.
- W3194086169 hasBestOaLocation W31940861691 @default.
- W3194086169 hasConcept C115961682 @default.
- W3194086169 hasConcept C124681953 @default.
- W3194086169 hasConcept C153180895 @default.
- W3194086169 hasConcept C154945302 @default.
- W3194086169 hasConcept C18903297 @default.
- W3194086169 hasConcept C2781238097 @default.