Matches in SemOpenAlex for { <https://semopenalex.org/work/W3106937659> ?p ?o ?g. }
- W3106937659 endingPage "595" @default.
- W3106937659 startingPage "578" @default.
- W3106937659 abstract "Deep neural networks are widely deployed on edge devices (e.g., for computer vision and speech recognition). Users either perform the inference locally (i.e., edge-based) or send the data to the cloud and run inference remotely (i.e., cloud-based). However, both solutions have their limitations: edge devices are heavily constrained by insufficient hardware resources and cannot afford to run large models; cloud servers, if not trustworthy, will raise serious privacy issues. In this paper, we mediate between the resource-constrained edge devices and the privacy-invasive cloud servers by introducing a novel privacy-preserving edge-cloud inference framework, DataMix. We off-load the majority of the computations to the cloud and leverage a pair of mixing and de-mixing operation, inspired by mixup, to protect the privacy of the data transmitted to the cloud. Our framework has three advantages. First, it is privacy-preserving as the mixing cannot be inverted without the user’s private mixing coefficients. Second, our framework is accuracy-preserving because our framework takes advantage of the space spanned by images, and we train the model in a mixing-aware manner to maintain accuracy. Third, our solution is efficient on the edge since the majority of the workload is delegated to the cloud, and our mixing and de-mixing processes introduce very few extra computations. Also, our framework introduces small communication overhead and maintains high hardware utilization on the cloud. Extensive experiments on multiple computer vision and speech recognition datasets demonstrate that our framework can greatly reduce the local computations on the edge (to fewer than 20% of FLOPs) with negligible loss of accuracy and no leakages of private information." @default.
- W3106937659 created "2020-12-07" @default.
- W3106937659 creator A5038230110 @default.
- W3106937659 creator A5040877128 @default.
- W3106937659 creator A5070926896 @default.
- W3106937659 creator A5073936760 @default.
- W3106937659 creator A5082308834 @default.
- W3106937659 date "2020-01-01" @default.
- W3106937659 modified "2023-10-14" @default.
- W3106937659 title "DataMix: Efficient Privacy-Preserving Edge-Cloud Inference" @default.
- W3106937659 cites W1488228643 @default.
- W3106937659 cites W1578378315 @default.
- W3106937659 cites W1834627138 @default.
- W3106937659 cites W1996901117 @default.
- W3106937659 cites W1999602050 @default.
- W3106937659 cites W2053637704 @default.
- W3106937659 cites W2103958416 @default.
- W3106937659 cites W2112941959 @default.
- W3106937659 cites W2120582309 @default.
- W3106937659 cites W2158551678 @default.
- W3106937659 cites W2159024459 @default.
- W3106937659 cites W2194775991 @default.
- W3106937659 cites W2473418344 @default.
- W3106937659 cites W2480709345 @default.
- W3106937659 cites W2526578496 @default.
- W3106937659 cites W2537024514 @default.
- W3106937659 cites W2605289356 @default.
- W3106937659 cites W2741234481 @default.
- W3106937659 cites W2743671162 @default.
- W3106937659 cites W2809414288 @default.
- W3106937659 cites W2883780447 @default.
- W3106937659 cites W2884985453 @default.
- W3106937659 cites W2886851211 @default.
- W3106937659 cites W2962697884 @default.
- W3106937659 cites W2962712356 @default.
- W3106937659 cites W2962786991 @default.
- W3106937659 cites W2962851801 @default.
- W3106937659 cites W2963073614 @default.
- W3106937659 cites W2963125010 @default.
- W3106937659 cites W2963163009 @default.
- W3106937659 cites W2963278610 @default.
- W3106937659 cites W2963363373 @default.
- W3106937659 cites W2977362526 @default.
- W3106937659 cites W2982083293 @default.
- W3106937659 cites W2982479999 @default.
- W3106937659 cites W2990152177 @default.
- W3106937659 cites W3035160371 @default.
- W3106937659 cites W4236440698 @default.
- W3106937659 cites W4248358572 @default.
- W3106937659 doi "https://doi.org/10.1007/978-3-030-58621-8_34" @default.
- W3106937659 hasPublicationYear "2020" @default.
- W3106937659 type Work @default.
- W3106937659 sameAs 3106937659 @default.
- W3106937659 citedByCount "10" @default.
- W3106937659 countsByYear W31069376592020 @default.
- W3106937659 countsByYear W31069376592021 @default.
- W3106937659 countsByYear W31069376592022 @default.
- W3106937659 countsByYear W31069376592023 @default.
- W3106937659 crossrefType "book-chapter" @default.
- W3106937659 hasAuthorship W3106937659A5038230110 @default.
- W3106937659 hasAuthorship W3106937659A5040877128 @default.
- W3106937659 hasAuthorship W3106937659A5070926896 @default.
- W3106937659 hasAuthorship W3106937659A5073936760 @default.
- W3106937659 hasAuthorship W3106937659A5082308834 @default.
- W3106937659 hasConcept C111919701 @default.
- W3106937659 hasConcept C11413529 @default.
- W3106937659 hasConcept C120314980 @default.
- W3106937659 hasConcept C121332964 @default.
- W3106937659 hasConcept C138236772 @default.
- W3106937659 hasConcept C138777275 @default.
- W3106937659 hasConcept C153083717 @default.
- W3106937659 hasConcept C154945302 @default.
- W3106937659 hasConcept C162307627 @default.
- W3106937659 hasConcept C2776214188 @default.
- W3106937659 hasConcept C2778456923 @default.
- W3106937659 hasConcept C2779960059 @default.
- W3106937659 hasConcept C31258907 @default.
- W3106937659 hasConcept C41008148 @default.
- W3106937659 hasConcept C45374587 @default.
- W3106937659 hasConcept C62520636 @default.
- W3106937659 hasConcept C79974875 @default.
- W3106937659 hasConcept C93996380 @default.
- W3106937659 hasConceptScore W3106937659C111919701 @default.
- W3106937659 hasConceptScore W3106937659C11413529 @default.
- W3106937659 hasConceptScore W3106937659C120314980 @default.
- W3106937659 hasConceptScore W3106937659C121332964 @default.
- W3106937659 hasConceptScore W3106937659C138236772 @default.
- W3106937659 hasConceptScore W3106937659C138777275 @default.
- W3106937659 hasConceptScore W3106937659C153083717 @default.
- W3106937659 hasConceptScore W3106937659C154945302 @default.
- W3106937659 hasConceptScore W3106937659C162307627 @default.
- W3106937659 hasConceptScore W3106937659C2776214188 @default.
- W3106937659 hasConceptScore W3106937659C2778456923 @default.
- W3106937659 hasConceptScore W3106937659C2779960059 @default.
- W3106937659 hasConceptScore W3106937659C31258907 @default.
- W3106937659 hasConceptScore W3106937659C41008148 @default.
- W3106937659 hasConceptScore W3106937659C45374587 @default.
- W3106937659 hasConceptScore W3106937659C62520636 @default.