Matches in SemOpenAlex for { <https://semopenalex.org/work/W4308065195> ?p ?o ?g. }
Showing items 1 to 64 of
64
with 100 items per page.
- W4308065195 abstract "Deep neural networks are discovered to be non-robust when attacked by imperceptible adversarial examples, which is dangerous for it applied into medical diagnostic system that requires high reliability. However, the defense methods that have good effect in natural images may not be suitable for medical diagnostic tasks. The preprocessing methods (e.g., random resizing, compression) may lead to the loss of the small lesions feature in the medical image. Retraining the network on the augmented data set is also not practical for medical models that have already been deployed online. Accordingly, it is necessary to design an easy-to-deploy and effective defense framework for medical diagnostic tasks. In this paper, we propose a Robust and Retrain-Less Diagnostic Framework for Medical pretrained models against adversarial attack (i.e., MedRDF). It acts on the inference time of the pertained medical model. Specifically, for each test image, MedRDF firstly creates a large number of noisy copies of it, and obtains the output labels of these copies from the pretrained medical diagnostic model. Then, based on the labels of these copies, MedRDF outputs the final robust diagnostic result by majority voting. In addition to the diagnostic result, MedRDF produces the Robust Metric (RM) as the confidence of the result. Therefore, it is convenient and reliable to utilize MedRDF to convert pre-trained non-robust diagnostic models into robust ones. The experimental results on COVID-19 and DermaMNIST datasets verify the effectiveness of our MedRDF in improving the robustness of medical diagnostic models." @default.
- W4308065195 created "2022-11-07" @default.
- W4308065195 creator A5018821033 @default.
- W4308065195 creator A5035701638 @default.
- W4308065195 creator A5041633536 @default.
- W4308065195 date "2021-11-29" @default.
- W4308065195 modified "2023-10-17" @default.
- W4308065195 title "MedRDF: A Robust and Retrain-Less Diagnostic Framework for Medical Pretrained Models Against Adversarial Attack" @default.
- W4308065195 doi "https://doi.org/10.48550/arxiv.2111.14564" @default.
- W4308065195 hasPublicationYear "2021" @default.
- W4308065195 type Work @default.
- W4308065195 citedByCount "0" @default.
- W4308065195 crossrefType "posted-content" @default.
- W4308065195 hasAuthorship W4308065195A5018821033 @default.
- W4308065195 hasAuthorship W4308065195A5035701638 @default.
- W4308065195 hasAuthorship W4308065195A5041633536 @default.
- W4308065195 hasBestOaLocation W43080651951 @default.
- W4308065195 hasConcept C104317684 @default.
- W4308065195 hasConcept C119857082 @default.
- W4308065195 hasConcept C124101348 @default.
- W4308065195 hasConcept C153180895 @default.
- W4308065195 hasConcept C154945302 @default.
- W4308065195 hasConcept C177264268 @default.
- W4308065195 hasConcept C185592680 @default.
- W4308065195 hasConcept C199360897 @default.
- W4308065195 hasConcept C2776214188 @default.
- W4308065195 hasConcept C31601959 @default.
- W4308065195 hasConcept C34736171 @default.
- W4308065195 hasConcept C37736160 @default.
- W4308065195 hasConcept C41008148 @default.
- W4308065195 hasConcept C55493867 @default.
- W4308065195 hasConcept C63479239 @default.
- W4308065195 hasConceptScore W4308065195C104317684 @default.
- W4308065195 hasConceptScore W4308065195C119857082 @default.
- W4308065195 hasConceptScore W4308065195C124101348 @default.
- W4308065195 hasConceptScore W4308065195C153180895 @default.
- W4308065195 hasConceptScore W4308065195C154945302 @default.
- W4308065195 hasConceptScore W4308065195C177264268 @default.
- W4308065195 hasConceptScore W4308065195C185592680 @default.
- W4308065195 hasConceptScore W4308065195C199360897 @default.
- W4308065195 hasConceptScore W4308065195C2776214188 @default.
- W4308065195 hasConceptScore W4308065195C31601959 @default.
- W4308065195 hasConceptScore W4308065195C34736171 @default.
- W4308065195 hasConceptScore W4308065195C37736160 @default.
- W4308065195 hasConceptScore W4308065195C41008148 @default.
- W4308065195 hasConceptScore W4308065195C55493867 @default.
- W4308065195 hasConceptScore W4308065195C63479239 @default.
- W4308065195 hasLocation W43080651951 @default.
- W4308065195 hasLocation W43080651952 @default.
- W4308065195 hasOpenAccess W4308065195 @default.
- W4308065195 hasPrimaryLocation W43080651951 @default.
- W4308065195 hasRelatedWork W2066259560 @default.
- W4308065195 hasRelatedWork W2380927352 @default.
- W4308065195 hasRelatedWork W2383487638 @default.
- W4308065195 hasRelatedWork W2391959412 @default.
- W4308065195 hasRelatedWork W4280534256 @default.
- W4308065195 hasRelatedWork W4307534228 @default.
- W4308065195 hasRelatedWork W4307654699 @default.
- W4308065195 hasRelatedWork W4311734044 @default.
- W4308065195 hasRelatedWork W4379255972 @default.
- W4308065195 hasRelatedWork W4383955378 @default.
- W4308065195 isParatext "false" @default.
- W4308065195 isRetracted "false" @default.
- W4308065195 workType "article" @default.