Matches in SemOpenAlex for { <https://semopenalex.org/work/W4383605000> ?p ?o ?g. }
Showing items 1 to 57 of
57
with 100 items per page.
- W4383605000 abstract "Generative Adversarial Networks (GANs) have demonstrated their ability to generate synthetic samples that match a target distribution. However, from a privacy perspective, using GANs as a proxy for data sharing is not a safe solution, as they tend to embed near-duplicates of real samples in the latent space. Recent works, inspired by k-anonymity principles, address this issue through sample aggregation in the latent space, with the drawback of reducing the dataset by a factor of k. Our work aims to mitigate this problem by proposing a latent space navigation strategy able to generate diverse synthetic samples that may support effective training of deep models, while addressing privacy concerns in a principled way. Our approach leverages an auxiliary identity classifier as a guide to non-linearly walk between points in the latent space, minimizing the risk of collision with near-duplicates of real samples. We empirically demonstrate that, given any random pair of points in the latent space, our walking strategy is safer than linear interpolation. We then test our path-finding strategy combined to k-same methods and demonstrate, on two benchmarks for tuberculosis and diabetic retinopathy classification, that training a model using samples generated by our approach mitigate drops in performance, while keeping privacy preservation." @default.
- W4383605000 created "2023-07-08" @default.
- W4383605000 creator A5003973109 @default.
- W4383605000 creator A5030188696 @default.
- W4383605000 creator A5031450889 @default.
- W4383605000 creator A5033378482 @default.
- W4383605000 creator A5075815307 @default.
- W4383605000 creator A5083136342 @default.
- W4383605000 date "2023-07-06" @default.
- W4383605000 modified "2023-10-14" @default.
- W4383605000 title "A Privacy-Preserving Walk in the Latent Space of Generative Models for Medical Applications" @default.
- W4383605000 doi "https://doi.org/10.48550/arxiv.2307.02984" @default.
- W4383605000 hasPublicationYear "2023" @default.
- W4383605000 type Work @default.
- W4383605000 citedByCount "0" @default.
- W4383605000 crossrefType "posted-content" @default.
- W4383605000 hasAuthorship W4383605000A5003973109 @default.
- W4383605000 hasAuthorship W4383605000A5030188696 @default.
- W4383605000 hasAuthorship W4383605000A5031450889 @default.
- W4383605000 hasAuthorship W4383605000A5033378482 @default.
- W4383605000 hasAuthorship W4383605000A5075815307 @default.
- W4383605000 hasAuthorship W4383605000A5083136342 @default.
- W4383605000 hasBestOaLocation W43836050001 @default.
- W4383605000 hasConcept C119857082 @default.
- W4383605000 hasConcept C124101348 @default.
- W4383605000 hasConcept C154945302 @default.
- W4383605000 hasConcept C167966045 @default.
- W4383605000 hasConcept C2776654903 @default.
- W4383605000 hasConcept C38652104 @default.
- W4383605000 hasConcept C39890363 @default.
- W4383605000 hasConcept C41008148 @default.
- W4383605000 hasConcept C95623464 @default.
- W4383605000 hasConceptScore W4383605000C119857082 @default.
- W4383605000 hasConceptScore W4383605000C124101348 @default.
- W4383605000 hasConceptScore W4383605000C154945302 @default.
- W4383605000 hasConceptScore W4383605000C167966045 @default.
- W4383605000 hasConceptScore W4383605000C2776654903 @default.
- W4383605000 hasConceptScore W4383605000C38652104 @default.
- W4383605000 hasConceptScore W4383605000C39890363 @default.
- W4383605000 hasConceptScore W4383605000C41008148 @default.
- W4383605000 hasConceptScore W4383605000C95623464 @default.
- W4383605000 hasLocation W43836050001 @default.
- W4383605000 hasOpenAccess W4383605000 @default.
- W4383605000 hasPrimaryLocation W43836050001 @default.
- W4383605000 hasRelatedWork W1576360539 @default.
- W4383605000 hasRelatedWork W2105080633 @default.
- W4383605000 hasRelatedWork W2166704235 @default.
- W4383605000 hasRelatedWork W2623427976 @default.
- W4383605000 hasRelatedWork W2805430782 @default.
- W4383605000 hasRelatedWork W2961085424 @default.
- W4383605000 hasRelatedWork W3212218375 @default.
- W4383605000 hasRelatedWork W4225729320 @default.
- W4383605000 hasRelatedWork W4310424780 @default.
- W4383605000 hasRelatedWork W4321153836 @default.
- W4383605000 isParatext "false" @default.
- W4383605000 isRetracted "false" @default.
- W4383605000 workType "article" @default.