Matches in SemOpenAlex for { <https://semopenalex.org/work/W4287753619> ?p ?o ?g. }
Showing items 1 to 77 of
77
with 100 items per page.
- W4287753619 abstract "Recent developments related to generative models have made it possible to generate diverse high-fidelity images. In particular, layout-to-image generation models have gained significant attention due to their capability to generate realistic complex images containing distinct objects. These models are generally conditioned on either semantic layouts or textual descriptions. However, unlike natural images, providing auxiliary information can be extremely hard in domains such as biomedical imaging and remote sensing. In this work, we propose a multi-object generation framework that can synthesize images with multiple objects without explicitly requiring their contextual information during the generation process. Based on a vector-quantized variational autoencoder (VQ-VAE) backbone, our model learns to preserve spatial coherency within an image as well as semantic coherency between the objects and the background through two powerful autoregressive priors: PixelSNAIL and LayoutPixelSNAIL. While the PixelSNAIL learns the distribution of the latent encodings of the VQ-VAE, the LayoutPixelSNAIL is used to specifically learn the semantic distribution of the objects. An implicit advantage of our approach is that the generated samples are accompanied by object-level annotations. We demonstrate how coherency and fidelity are preserved with our method through experiments on the Multi-MNIST and CLEVR datasets; thereby outperforming state-of-the-art multi-object generative methods. The efficacy of our approach is demonstrated through application on medical imaging datasets, where we show that augmenting the training set with generated samples using our approach improves the performance of existing models." @default.
- W4287753619 created "2022-07-26" @default.
- W4287753619 creator A5003986814 @default.
- W4287753619 creator A5055224011 @default.
- W4287753619 creator A5078127969 @default.
- W4287753619 date "2020-06-22" @default.
- W4287753619 modified "2023-09-27" @default.
- W4287753619 title "Generating Annotated High-Fidelity Images Containing Multiple Coherent Objects" @default.
- W4287753619 doi "https://doi.org/10.48550/arxiv.2006.12150" @default.
- W4287753619 hasPublicationYear "2020" @default.
- W4287753619 type Work @default.
- W4287753619 citedByCount "0" @default.
- W4287753619 crossrefType "posted-content" @default.
- W4287753619 hasAuthorship W4287753619A5003986814 @default.
- W4287753619 hasAuthorship W4287753619A5055224011 @default.
- W4287753619 hasAuthorship W4287753619A5078127969 @default.
- W4287753619 hasBestOaLocation W42877536191 @default.
- W4287753619 hasConcept C101738243 @default.
- W4287753619 hasConcept C107673813 @default.
- W4287753619 hasConcept C108583219 @default.
- W4287753619 hasConcept C111919701 @default.
- W4287753619 hasConcept C113364801 @default.
- W4287753619 hasConcept C115961682 @default.
- W4287753619 hasConcept C119599485 @default.
- W4287753619 hasConcept C119857082 @default.
- W4287753619 hasConcept C127413603 @default.
- W4287753619 hasConcept C153180895 @default.
- W4287753619 hasConcept C154945302 @default.
- W4287753619 hasConcept C167966045 @default.
- W4287753619 hasConcept C177264268 @default.
- W4287753619 hasConcept C177769412 @default.
- W4287753619 hasConcept C190502265 @default.
- W4287753619 hasConcept C199360897 @default.
- W4287753619 hasConcept C2776459999 @default.
- W4287753619 hasConcept C2781238097 @default.
- W4287753619 hasConcept C39890363 @default.
- W4287753619 hasConcept C41008148 @default.
- W4287753619 hasConcept C76155785 @default.
- W4287753619 hasConcept C98045186 @default.
- W4287753619 hasConceptScore W4287753619C101738243 @default.
- W4287753619 hasConceptScore W4287753619C107673813 @default.
- W4287753619 hasConceptScore W4287753619C108583219 @default.
- W4287753619 hasConceptScore W4287753619C111919701 @default.
- W4287753619 hasConceptScore W4287753619C113364801 @default.
- W4287753619 hasConceptScore W4287753619C115961682 @default.
- W4287753619 hasConceptScore W4287753619C119599485 @default.
- W4287753619 hasConceptScore W4287753619C119857082 @default.
- W4287753619 hasConceptScore W4287753619C127413603 @default.
- W4287753619 hasConceptScore W4287753619C153180895 @default.
- W4287753619 hasConceptScore W4287753619C154945302 @default.
- W4287753619 hasConceptScore W4287753619C167966045 @default.
- W4287753619 hasConceptScore W4287753619C177264268 @default.
- W4287753619 hasConceptScore W4287753619C177769412 @default.
- W4287753619 hasConceptScore W4287753619C190502265 @default.
- W4287753619 hasConceptScore W4287753619C199360897 @default.
- W4287753619 hasConceptScore W4287753619C2776459999 @default.
- W4287753619 hasConceptScore W4287753619C2781238097 @default.
- W4287753619 hasConceptScore W4287753619C39890363 @default.
- W4287753619 hasConceptScore W4287753619C41008148 @default.
- W4287753619 hasConceptScore W4287753619C76155785 @default.
- W4287753619 hasConceptScore W4287753619C98045186 @default.
- W4287753619 hasLocation W42877536191 @default.
- W4287753619 hasOpenAccess W4287753619 @default.
- W4287753619 hasPrimaryLocation W42877536191 @default.
- W4287753619 hasRelatedWork W2578760059 @default.
- W4287753619 hasRelatedWork W2891962740 @default.
- W4287753619 hasRelatedWork W2904927891 @default.
- W4287753619 hasRelatedWork W2905104183 @default.
- W4287753619 hasRelatedWork W3037763638 @default.
- W4287753619 hasRelatedWork W3152912388 @default.
- W4287753619 hasRelatedWork W4287753619 @default.
- W4287753619 hasRelatedWork W4293582597 @default.
- W4287753619 hasRelatedWork W4315880743 @default.
- W4287753619 hasRelatedWork W4321855132 @default.
- W4287753619 isParatext "false" @default.
- W4287753619 isRetracted "false" @default.
- W4287753619 workType "article" @default.