Matches in SemOpenAlex for { <https://semopenalex.org/work/W2982620140> ?p ?o ?g. }
- W2982620140 abstract "Generative models produce realistic objects in many domains, including text, image, video, and audio synthesis. Most popular models---Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs)---usually employ a standard Gaussian distribution as a prior. Previous works show that the richer family of prior distributions may help to avoid the mode collapse problem in GANs and to improve the evidence lower bound in VAEs. We propose a new family of prior distributions---Tensor Ring Induced Prior (TRIP)---that packs an exponential number of Gaussians into a high-dimensional lattice with a relatively small number of parameters. We show that these priors improve Frechet Inception Distance for GANs and Evidence Lower Bound for VAEs. We also study generative models with TRIP in the conditional generation setup with missing conditions. Altogether, we propose a novel plug-and-play framework for generative models that can be utilized in any GAN and VAE-like architectures." @default.
- W2982620140 created "2019-11-08" @default.
- W2982620140 creator A5021794857 @default.
- W2982620140 creator A5028940329 @default.
- W2982620140 creator A5045908724 @default.
- W2982620140 creator A5061298160 @default.
- W2982620140 date "2019-10-29" @default.
- W2982620140 modified "2023-09-27" @default.
- W2982620140 title "A Prior of a Googol Gaussians: a Tensor Ring Induced Prior for Generative Models." @default.
- W2982620140 cites W1798945469 @default.
- W2982620140 cites W1834627138 @default.
- W2982620140 cites W1959608418 @default.
- W2982620140 cites W1993482030 @default.
- W2982620140 cites W2099471712 @default.
- W2982620140 cites W2119717200 @default.
- W2982620140 cites W2431890537 @default.
- W2982620140 cites W2559813832 @default.
- W2982620140 cites W2560512785 @default.
- W2982620140 cites W2587284713 @default.
- W2982620140 cites W2623550831 @default.
- W2982620140 cites W2769810959 @default.
- W2982620140 cites W2889205371 @default.
- W2982620140 cites W2891868449 @default.
- W2982620140 cites W2904303242 @default.
- W2982620140 cites W2962695743 @default.
- W2982620140 cites W2962738009 @default.
- W2982620140 cites W2962760235 @default.
- W2982620140 cites W2962879692 @default.
- W2982620140 cites W2962897886 @default.
- W2982620140 cites W2963044468 @default.
- W2982620140 cites W2963082441 @default.
- W2982620140 cites W2963090522 @default.
- W2982620140 cites W2963275229 @default.
- W2982620140 cites W2963650472 @default.
- W2982620140 cites W2963691546 @default.
- W2982620140 cites W2963799213 @default.
- W2982620140 cites W2963836885 @default.
- W2982620140 cites W2963948515 @default.
- W2982620140 cites W2963981733 @default.
- W2982620140 cites W2964020555 @default.
- W2982620140 cites W2964121744 @default.
- W2982620140 cites W2971690404 @default.
- W2982620140 hasPublicationYear "2019" @default.
- W2982620140 type Work @default.
- W2982620140 sameAs 2982620140 @default.
- W2982620140 citedByCount "0" @default.
- W2982620140 crossrefType "posted-content" @default.
- W2982620140 hasAuthorship W2982620140A5021794857 @default.
- W2982620140 hasAuthorship W2982620140A5028940329 @default.
- W2982620140 hasAuthorship W2982620140A5045908724 @default.
- W2982620140 hasAuthorship W2982620140A5061298160 @default.
- W2982620140 hasConcept C107673813 @default.
- W2982620140 hasConcept C11413529 @default.
- W2982620140 hasConcept C119857082 @default.
- W2982620140 hasConcept C121332964 @default.
- W2982620140 hasConcept C134306372 @default.
- W2982620140 hasConcept C153180895 @default.
- W2982620140 hasConcept C154945302 @default.
- W2982620140 hasConcept C163716315 @default.
- W2982620140 hasConcept C167966045 @default.
- W2982620140 hasConcept C177769412 @default.
- W2982620140 hasConcept C33923547 @default.
- W2982620140 hasConcept C39890363 @default.
- W2982620140 hasConcept C41008148 @default.
- W2982620140 hasConcept C55974624 @default.
- W2982620140 hasConcept C61224824 @default.
- W2982620140 hasConcept C62520636 @default.
- W2982620140 hasConcept C77553402 @default.
- W2982620140 hasConceptScore W2982620140C107673813 @default.
- W2982620140 hasConceptScore W2982620140C11413529 @default.
- W2982620140 hasConceptScore W2982620140C119857082 @default.
- W2982620140 hasConceptScore W2982620140C121332964 @default.
- W2982620140 hasConceptScore W2982620140C134306372 @default.
- W2982620140 hasConceptScore W2982620140C153180895 @default.
- W2982620140 hasConceptScore W2982620140C154945302 @default.
- W2982620140 hasConceptScore W2982620140C163716315 @default.
- W2982620140 hasConceptScore W2982620140C167966045 @default.
- W2982620140 hasConceptScore W2982620140C177769412 @default.
- W2982620140 hasConceptScore W2982620140C33923547 @default.
- W2982620140 hasConceptScore W2982620140C39890363 @default.
- W2982620140 hasConceptScore W2982620140C41008148 @default.
- W2982620140 hasConceptScore W2982620140C55974624 @default.
- W2982620140 hasConceptScore W2982620140C61224824 @default.
- W2982620140 hasConceptScore W2982620140C62520636 @default.
- W2982620140 hasConceptScore W2982620140C77553402 @default.
- W2982620140 hasLocation W29826201401 @default.
- W2982620140 hasOpenAccess W2982620140 @default.
- W2982620140 hasPrimaryLocation W29826201401 @default.
- W2982620140 hasRelatedWork W2152498207 @default.
- W2982620140 hasRelatedWork W2159528849 @default.
- W2982620140 hasRelatedWork W2209787065 @default.
- W2982620140 hasRelatedWork W2271522117 @default.
- W2982620140 hasRelatedWork W2403392501 @default.
- W2982620140 hasRelatedWork W2421157170 @default.
- W2982620140 hasRelatedWork W2532488433 @default.
- W2982620140 hasRelatedWork W2727959328 @default.
- W2982620140 hasRelatedWork W2891413839 @default.
- W2982620140 hasRelatedWork W2951999424 @default.
- W2982620140 hasRelatedWork W2970754141 @default.
- W2982620140 hasRelatedWork W2996123903 @default.