Matches in SemOpenAlex for { <https://semopenalex.org/work/W4301003206> ?p ?o ?g. }
Showing items 1 to 66 of
66
with 100 items per page.
- W4301003206 abstract "Deep neural networks (DNNs) have recently been achieving state-of-the-art performance on a variety of pattern-recognition tasks, most notably visual classification problems. Given that DNNs are now able to classify objects in images with near-human-level performance, questions naturally arise as to what differences remain between computer and human vision. A recent study revealed that changing an image (e.g. of a lion) in a way imperceptible to humans can cause a DNN to label the image as something else entirely (e.g. mislabeling a lion a library). Here we show a related result: it is easy to produce images that are completely unrecognizable to humans, but that state-of-the-art DNNs believe to be recognizable objects with 99.99% confidence (e.g. labeling with certainty that white noise static is a lion). Specifically, we take convolutional neural networks trained to perform well on either the ImageNet or MNIST datasets and then find images with evolutionary algorithms or gradient ascent that DNNs label with high confidence as belonging to each dataset class. It is possible to produce images totally unrecognizable to human eyes that DNNs believe with near certainty are familiar objects, which we call fooling images (more generally, fooling examples). Our results shed light on interesting differences between human vision and current DNNs, and raise questions about the generality of DNN computer vision." @default.
- W4301003206 created "2022-10-04" @default.
- W4301003206 creator A5006797018 @default.
- W4301003206 creator A5045605838 @default.
- W4301003206 creator A5049659753 @default.
- W4301003206 date "2014-12-05" @default.
- W4301003206 modified "2023-10-03" @default.
- W4301003206 title "Deep Neural Networks are Easily Fooled: High Confidence Predictions for Unrecognizable Images" @default.
- W4301003206 doi "https://doi.org/10.48550/arxiv.1412.1897" @default.
- W4301003206 hasPublicationYear "2014" @default.
- W4301003206 type Work @default.
- W4301003206 citedByCount "4" @default.
- W4301003206 countsByYear W43010032062022 @default.
- W4301003206 countsByYear W43010032062023 @default.
- W4301003206 crossrefType "posted-content" @default.
- W4301003206 hasAuthorship W4301003206A5006797018 @default.
- W4301003206 hasAuthorship W4301003206A5045605838 @default.
- W4301003206 hasAuthorship W4301003206A5049659753 @default.
- W4301003206 hasBestOaLocation W43010032061 @default.
- W4301003206 hasConcept C108583219 @default.
- W4301003206 hasConcept C115961682 @default.
- W4301003206 hasConcept C119857082 @default.
- W4301003206 hasConcept C136197465 @default.
- W4301003206 hasConcept C153180895 @default.
- W4301003206 hasConcept C154945302 @default.
- W4301003206 hasConcept C15744967 @default.
- W4301003206 hasConcept C190502265 @default.
- W4301003206 hasConcept C2780767217 @default.
- W4301003206 hasConcept C2781238097 @default.
- W4301003206 hasConcept C2984842247 @default.
- W4301003206 hasConcept C41008148 @default.
- W4301003206 hasConcept C50644808 @default.
- W4301003206 hasConcept C542102704 @default.
- W4301003206 hasConcept C81363708 @default.
- W4301003206 hasConceptScore W4301003206C108583219 @default.
- W4301003206 hasConceptScore W4301003206C115961682 @default.
- W4301003206 hasConceptScore W4301003206C119857082 @default.
- W4301003206 hasConceptScore W4301003206C136197465 @default.
- W4301003206 hasConceptScore W4301003206C153180895 @default.
- W4301003206 hasConceptScore W4301003206C154945302 @default.
- W4301003206 hasConceptScore W4301003206C15744967 @default.
- W4301003206 hasConceptScore W4301003206C190502265 @default.
- W4301003206 hasConceptScore W4301003206C2780767217 @default.
- W4301003206 hasConceptScore W4301003206C2781238097 @default.
- W4301003206 hasConceptScore W4301003206C2984842247 @default.
- W4301003206 hasConceptScore W4301003206C41008148 @default.
- W4301003206 hasConceptScore W4301003206C50644808 @default.
- W4301003206 hasConceptScore W4301003206C542102704 @default.
- W4301003206 hasConceptScore W4301003206C81363708 @default.
- W4301003206 hasLocation W43010032061 @default.
- W4301003206 hasLocation W43010032062 @default.
- W4301003206 hasOpenAccess W4301003206 @default.
- W4301003206 hasPrimaryLocation W43010032061 @default.
- W4301003206 hasRelatedWork W2597787948 @default.
- W4301003206 hasRelatedWork W2732542196 @default.
- W4301003206 hasRelatedWork W2738221750 @default.
- W4301003206 hasRelatedWork W2904174853 @default.
- W4301003206 hasRelatedWork W2947175736 @default.
- W4301003206 hasRelatedWork W2951786554 @default.
- W4301003206 hasRelatedWork W2978290780 @default.
- W4301003206 hasRelatedWork W3156786002 @default.
- W4301003206 hasRelatedWork W4309224979 @default.
- W4301003206 hasRelatedWork W4311257506 @default.
- W4301003206 isParatext "false" @default.
- W4301003206 isRetracted "false" @default.
- W4301003206 workType "article" @default.