Matches in SemOpenAlex for { <https://semopenalex.org/work/W2558647090> ?p ?o ?g. }
- W2558647090 abstract "Variational autoencoders (VAEs), that are built upon deep neural networks have emerged as popular generative models in computer vision. Most of the work towards improving variational autoencoders has focused mainly on making the approximations to the posterior flexible and accurate, leading to tremendous progress. However, there have been limited efforts to replace pixel-wise reconstruction, which have known shortcomings. In this work, we use real-valued non-volume preserving transformations (real NVP) to exactly compute the conditional likelihood of the data given the latent distribution. We show that a simple VAE with this form of reconstruction is competitive with complicated VAE structures, on image modeling tasks. As part of our model, we develop powerful conditional coupling layers that enable real NVP to learn with fewer intermediate layers." @default.
- W2558647090 created "2016-12-08" @default.
- W2558647090 creator A5010292297 @default.
- W2558647090 creator A5018019178 @default.
- W2558647090 date "2016-11-16" @default.
- W2558647090 modified "2023-09-27" @default.
- W2558647090 title "Deep Variational Inference Without Pixel-Wise Reconstruction" @default.
- W2558647090 cites W1691728462 @default.
- W2558647090 cites W1779483307 @default.
- W2558647090 cites W1834627138 @default.
- W2558647090 cites W1906598733 @default.
- W2558647090 cites W1909320841 @default.
- W2558647090 cites W2194775991 @default.
- W2558647090 cites W2202109488 @default.
- W2558647090 cites W2248556341 @default.
- W2558647090 cites W2409550820 @default.
- W2558647090 cites W2469894155 @default.
- W2558647090 cites W2949416428 @default.
- W2558647090 cites W2951004968 @default.
- W2558647090 cites W2953318193 @default.
- W2558647090 cites W2962741254 @default.
- W2558647090 cites W2963143316 @default.
- W2558647090 cites W2963147844 @default.
- W2558647090 cites W2963730200 @default.
- W2558647090 cites W2964121744 @default.
- W2558647090 cites W3118608800 @default.
- W2558647090 hasPublicationYear "2016" @default.
- W2558647090 type Work @default.
- W2558647090 sameAs 2558647090 @default.
- W2558647090 citedByCount "6" @default.
- W2558647090 countsByYear W25586470902019 @default.
- W2558647090 countsByYear W25586470902020 @default.
- W2558647090 countsByYear W25586470902021 @default.
- W2558647090 crossrefType "posted-content" @default.
- W2558647090 hasAuthorship W2558647090A5010292297 @default.
- W2558647090 hasAuthorship W2558647090A5018019178 @default.
- W2558647090 hasConcept C105795698 @default.
- W2558647090 hasConcept C108583219 @default.
- W2558647090 hasConcept C111472728 @default.
- W2558647090 hasConcept C11413529 @default.
- W2558647090 hasConcept C115961682 @default.
- W2558647090 hasConcept C119857082 @default.
- W2558647090 hasConcept C138885662 @default.
- W2558647090 hasConcept C153180895 @default.
- W2558647090 hasConcept C154945302 @default.
- W2558647090 hasConcept C160633673 @default.
- W2558647090 hasConcept C167966045 @default.
- W2558647090 hasConcept C2776214188 @default.
- W2558647090 hasConcept C2780586882 @default.
- W2558647090 hasConcept C2984842247 @default.
- W2558647090 hasConcept C33923547 @default.
- W2558647090 hasConcept C39890363 @default.
- W2558647090 hasConcept C41008148 @default.
- W2558647090 hasConcept C43555835 @default.
- W2558647090 hasConcept C50644808 @default.
- W2558647090 hasConceptScore W2558647090C105795698 @default.
- W2558647090 hasConceptScore W2558647090C108583219 @default.
- W2558647090 hasConceptScore W2558647090C111472728 @default.
- W2558647090 hasConceptScore W2558647090C11413529 @default.
- W2558647090 hasConceptScore W2558647090C115961682 @default.
- W2558647090 hasConceptScore W2558647090C119857082 @default.
- W2558647090 hasConceptScore W2558647090C138885662 @default.
- W2558647090 hasConceptScore W2558647090C153180895 @default.
- W2558647090 hasConceptScore W2558647090C154945302 @default.
- W2558647090 hasConceptScore W2558647090C160633673 @default.
- W2558647090 hasConceptScore W2558647090C167966045 @default.
- W2558647090 hasConceptScore W2558647090C2776214188 @default.
- W2558647090 hasConceptScore W2558647090C2780586882 @default.
- W2558647090 hasConceptScore W2558647090C2984842247 @default.
- W2558647090 hasConceptScore W2558647090C33923547 @default.
- W2558647090 hasConceptScore W2558647090C39890363 @default.
- W2558647090 hasConceptScore W2558647090C41008148 @default.
- W2558647090 hasConceptScore W2558647090C43555835 @default.
- W2558647090 hasConceptScore W2558647090C50644808 @default.
- W2558647090 hasLocation W25586470901 @default.
- W2558647090 hasOpenAccess W2558647090 @default.
- W2558647090 hasPrimaryLocation W25586470901 @default.
- W2558647090 hasRelatedWork W2267126114 @default.
- W2558647090 hasRelatedWork W2605145092 @default.
- W2558647090 hasRelatedWork W2788766568 @default.
- W2558647090 hasRelatedWork W2803580049 @default.
- W2558647090 hasRelatedWork W2949301738 @default.
- W2558647090 hasRelatedWork W2957120403 @default.
- W2558647090 hasRelatedWork W2962897886 @default.
- W2558647090 hasRelatedWork W2963090522 @default.
- W2558647090 hasRelatedWork W2963139417 @default.
- W2558647090 hasRelatedWork W2963864522 @default.
- W2558647090 hasRelatedWork W2995639611 @default.
- W2558647090 hasRelatedWork W2998280611 @default.
- W2558647090 hasRelatedWork W3034822012 @default.
- W2558647090 hasRelatedWork W3102025760 @default.
- W2558647090 hasRelatedWork W3118294324 @default.
- W2558647090 hasRelatedWork W3122499440 @default.
- W2558647090 hasRelatedWork W3123371847 @default.
- W2558647090 hasRelatedWork W3124839307 @default.
- W2558647090 hasRelatedWork W3128448487 @default.
- W2558647090 hasRelatedWork W3178129852 @default.
- W2558647090 isParatext "false" @default.
- W2558647090 isRetracted "false" @default.
- W2558647090 magId "2558647090" @default.