Matches in SemOpenAlex for { <https://semopenalex.org/work/W2896651608> ?p ?o ?g. }
Showing items 1 to 62 of
62
with 100 items per page.
- W2896651608 abstract "Deep generative models are powerful but difficult to train due to its instability, saturation problem and high dimensional data distribution. This paper introduces a game theory framework with Wasserstein metric to train generative models, in which the unknown data distribution is learned by dynamically optimizing the worst-case payoff. In the game, two types of players work on opposite objectives to solve a minimax problem. The defenders explore the Wasserstein neighborhood of real data to generate a set of hard samples which have the maximum distance from the model distribution. The attackers update the model to fit for the hard set so as to minimize the discrepancy between model and data distributions. Instead of Kullback-Leibler divergence, we use Wasserstein distance to measure the similarity between distributions. The Wasserstein metric is a true distance with better topology in the parameter space, which improves the stability of training. We provide practical algorithms to train deep generative models, in which an encoder network is designed to learn the feature vector of the high dimensional data. The algorithm is tested on CelebA human face dataset and compared with the state-of-the-art generative models. Performance evaluation shows the training process is stable and converges fast. Our model can produce visual pleasing images which are closer to the real distribution in terms of Wasserstein distance." @default.
- W2896651608 created "2018-10-26" @default.
- W2896651608 creator A5028795816 @default.
- W2896651608 creator A5073493112 @default.
- W2896651608 date "2018-07-01" @default.
- W2896651608 modified "2023-09-26" @default.
- W2896651608 title "Distributionally Robust Games: Wasserstein Metric" @default.
- W2896651608 cites W1834627138 @default.
- W2896651608 cites W1953936588 @default.
- W2896651608 cites W1985077192 @default.
- W2896651608 cites W1990275757 @default.
- W2896651608 cites W2003447360 @default.
- W2896651608 cites W2064630666 @default.
- W2896651608 cites W2066111511 @default.
- W2896651608 cites W2131116400 @default.
- W2896651608 cites W2136922672 @default.
- W2896651608 cites W2755863105 @default.
- W2896651608 cites W4233762729 @default.
- W2896651608 doi "https://doi.org/10.1109/ijcnn.2018.8489636" @default.
- W2896651608 hasPublicationYear "2018" @default.
- W2896651608 type Work @default.
- W2896651608 sameAs 2896651608 @default.
- W2896651608 citedByCount "7" @default.
- W2896651608 countsByYear W28966516082020 @default.
- W2896651608 countsByYear W28966516082021 @default.
- W2896651608 countsByYear W28966516082022 @default.
- W2896651608 crossrefType "proceedings-article" @default.
- W2896651608 hasAuthorship W2896651608A5028795816 @default.
- W2896651608 hasAuthorship W2896651608A5073493112 @default.
- W2896651608 hasConcept C126255220 @default.
- W2896651608 hasConcept C127413603 @default.
- W2896651608 hasConcept C176217482 @default.
- W2896651608 hasConcept C21547014 @default.
- W2896651608 hasConcept C2777634741 @default.
- W2896651608 hasConcept C28826006 @default.
- W2896651608 hasConcept C33923547 @default.
- W2896651608 hasConcept C41008148 @default.
- W2896651608 hasConceptScore W2896651608C126255220 @default.
- W2896651608 hasConceptScore W2896651608C127413603 @default.
- W2896651608 hasConceptScore W2896651608C176217482 @default.
- W2896651608 hasConceptScore W2896651608C21547014 @default.
- W2896651608 hasConceptScore W2896651608C2777634741 @default.
- W2896651608 hasConceptScore W2896651608C28826006 @default.
- W2896651608 hasConceptScore W2896651608C33923547 @default.
- W2896651608 hasConceptScore W2896651608C41008148 @default.
- W2896651608 hasLocation W28966516081 @default.
- W2896651608 hasOpenAccess W2896651608 @default.
- W2896651608 hasPrimaryLocation W28966516081 @default.
- W2896651608 hasRelatedWork W1968971433 @default.
- W2896651608 hasRelatedWork W2054697996 @default.
- W2896651608 hasRelatedWork W2522450843 @default.
- W2896651608 hasRelatedWork W266446692 @default.
- W2896651608 hasRelatedWork W2795913521 @default.
- W2896651608 hasRelatedWork W2910630640 @default.
- W2896651608 hasRelatedWork W2922216937 @default.
- W2896651608 hasRelatedWork W2964267938 @default.
- W2896651608 hasRelatedWork W2991626910 @default.
- W2896651608 hasRelatedWork W3014552397 @default.
- W2896651608 isParatext "false" @default.
- W2896651608 isRetracted "false" @default.
- W2896651608 magId "2896651608" @default.
- W2896651608 workType "article" @default.