FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3012402011> ?p ?o ?g. }

• W3012402011 abstract "Deep generative models are often used for human motion prediction as they are able to model multi-modal data distributions and characterize diverse human behavior. While much care has been taken into designing and learning deep generative models, how to efficiently produce diverse samples from a deep generative model after it has been trained is still an under-explored problem. To obtain samples from a pretrained generative model, most existing generative human motion prediction methods draw a set of independent Gaussian latent codes and convert them to motion samples. Clearly, this random sampling strategy is not guaranteed to produce diverse samples for two reasons: (1) The independent sampling cannot force the samples to be diverse; (2) The sampling is based solely on likelihood which may only produce samples that correspond to the major modes of the data distribution. To address these problems, we propose a novel sampling method, Diversifying Latent Flows (DLow), to produce a diverse set of samples from a pretrained deep generative model. Unlike random (independent) sampling, the proposed DLow sampling method samples a single random variable and then maps it with a set of learnable mapping functions to a set of correlated latent codes. The correlated latent codes are then decoded into a set of correlated samples. During training, DLow uses a diversity-promoting prior over samples as an objective to optimize the latent mappings to improve sample diversity. The design of the prior is highly flexible and can be customized to generate diverse motions with common features (e.g., similar leg motion but diverse upper-body motion). Our experiments demonstrate that DLow outperforms state-of-the-art baseline methods in terms of sample diversity and accuracy. Our code is released on the project page: https://www.ye-yuan.com/dlow." @default.
• W3012402011 created "2020-03-23" @default.
• W3012402011 creator A5037322163 @default.
• W3012402011 creator A5081199756 @default.
• W3012402011 date "2020-03-18" @default.
• W3012402011 modified "2023-09-24" @default.
• W3012402011 title "DLow: Diversifying Latent Flows for Diverse Human Motion Prediction" @default.
• W3012402011 cites W1516111018 @default.
• W3012402011 cites W1522301498 @default.
• W3012402011 cites W1603514996 @default.
• W3012402011 cites W1685464609 @default.
• W3012402011 cites W1735317348 @default.
• W3012402011 cites W1913984289 @default.
• W3012402011 cites W1924770834 @default.
• W3012402011 cites W2006672084 @default.
• W3012402011 cites W2012856192 @default.
• W3012402011 cites W2068782468 @default.
• W3012402011 cites W2092720120 @default.
• W3012402011 cites W2099333815 @default.
• W3012402011 cites W2099471712 @default.
• W3012402011 cites W2115857089 @default.
• W3012402011 cites W2144481990 @default.
• W3012402011 cites W2148985659 @default.
• W3012402011 cites W2158520534 @default.
• W3012402011 cites W2187580259 @default.
• W3012402011 cites W2343568200 @default.
• W3012402011 cites W2424778531 @default.
• W3012402011 cites W2556467266 @default.
• W3012402011 cites W2585630030 @default.
• W3012402011 cites W2594167370 @default.
• W3012402011 cites W2606517404 @default.
• W3012402011 cites W2606862842 @default.
• W3012402011 cites W2606884824 @default.
• W3012402011 cites W2607296803 @default.
• W3012402011 cites W2612706635 @default.
• W3012402011 cites W2622563070 @default.
• W3012402011 cites W2623550831 @default.
• W3012402011 cites W2748386424 @default.
• W3012402011 cites W2753738274 @default.
• W3012402011 cites W2772328249 @default.
• W3012402011 cites W2785734416 @default.
• W3012402011 cites W2788865504 @default.
• W3012402011 cites W2793483732 @default.
• W3012402011 cites W2795089319 @default.
• W3012402011 cites W2798670545 @default.
• W3012402011 cites W2887738788 @default.
• W3012402011 cites W2894978157 @default.
• W3012402011 cites W2902149428 @default.
• W3012402011 cites W2904383899 @default.
• W3012402011 cites W2949748686 @default.
• W3012402011 cites W2950304053 @default.
• W3012402011 cites W2951004968 @default.
• W3012402011 cites W2951670304 @default.
• W3012402011 cites W2962101532 @default.
• W3012402011 cites W2962879692 @default.
• W3012402011 cites W2962896489 @default.
• W3012402011 cites W2962916650 @default.
• W3012402011 cites W2962945412 @default.
• W3012402011 cites W2963001155 @default.
• W3012402011 cites W2963065614 @default.
• W3012402011 cites W2963142510 @default.
• W3012402011 cites W2963165299 @default.
• W3012402011 cites W2963226019 @default.
• W3012402011 cites W2963324990 @default.
• W3012402011 cites W2963389355 @default.
• W3012402011 cites W2963669520 @default.
• W3012402011 cites W2963746531 @default.
• W3012402011 cites W2963809789 @default.
• W3012402011 cites W2964000524 @default.
• W3012402011 cites W2964203186 @default.
• W3012402011 cites W2964468534 @default.
• W3012402011 cites W2982573856 @default.
• W3012402011 cites W2983925976 @default.
• W3012402011 cites W2991332406 @default.
• W3012402011 cites W2993797559 @default.
• W3012402011 cites W299440670 @default.
• W3012402011 cites W3004407381 @default.
• W3012402011 cites W3009150298 @default.
• W3012402011 cites W3011353537 @default.
• W3012402011 cites W3034938233 @default.
• W3012402011 cites W3035427621 @default.
• W3012402011 cites W3035613220 @default.
• W3012402011 cites W3103014337 @default.
• W3012402011 cites W3146927829 @default.
• W3012402011 doi "https://doi.org/10.48550/arxiv.2003.08386" @default.
• W3012402011 hasPublicationYear "2020" @default.
• W3012402011 type Work @default.
• W3012402011 sameAs 3012402011 @default.
• W3012402011 citedByCount "2" @default.
• W3012402011 countsByYear W30124020112019 @default.
• W3012402011 countsByYear W30124020112020 @default.
• W3012402011 crossrefType "posted-content" @default.
• W3012402011 hasAuthorship W3012402011A5037322163 @default.
• W3012402011 hasAuthorship W3012402011A5081199756 @default.
• W3012402011 hasBestOaLocation W30124020111 @default.
• W3012402011 hasConcept C104114177 @default.
• W3012402011 hasConcept C106131492 @default.
• W3012402011 hasConcept C119857082 @default.
• W3012402011 hasConcept C140779682 @default.
• W3012402011 hasConcept C153180895 @default.

• next