Matches in SemOpenAlex for { <https://semopenalex.org/work/W2779770929> ?p ?o ?g. }
Showing items 1 to 95 of
95
with 100 items per page.
- W2779770929 abstract "This work shows that it is possible to fool/attack recent state-of-the-art face detectors which are based on the single-stage networks. Successfully attacking face detectors could be a serious malware vulnerability when deploying a smart surveillance system utilizing face detectors. We show that existing adversarial perturbation methods are not effective to perform such an attack, especially when there are multiple faces in the input image. This is because the adversarial perturbation specifically generated for one face may disrupt the adversarial perturbation for another face. In this paper, we call this problem the Instance Perturbation Interference (IPI) problem. This IPI problem is addressed by studying the relationship between the deep neural network receptive field and the adversarial perturbation. As such, we propose the Localized Instance Perturbation (LIP) that uses adversarial perturbation constrained to the Effective Receptive Field (ERF) of a target to perform the attack. Experiment results show the LIP method massively outperforms existing adversarial perturbation generation methods -- often by a factor of 2 to 10." @default.
- W2779770929 created "2018-01-05" @default.
- W2779770929 creator A5020806354 @default.
- W2779770929 creator A5058539660 @default.
- W2779770929 creator A5063048878 @default.
- W2779770929 creator A5089800102 @default.
- W2779770929 date "2017-12-21" @default.
- W2779770929 modified "2023-09-24" @default.
- W2779770929 title "Using LIP to Gloss Over Faces in Single-Stage Face Detection Networks" @default.
- W2779770929 cites W1686810756 @default.
- W2779770929 cites W182571476 @default.
- W2779770929 cites W1903029394 @default.
- W2779770929 cites W1932198206 @default.
- W2779770929 cites W1934410531 @default.
- W2779770929 cites W204612701 @default.
- W2779770929 cites W2164598857 @default.
- W2779770929 cites W2176613063 @default.
- W2779770929 cites W2194775991 @default.
- W2779770929 cites W2243397390 @default.
- W2779770929 cites W2282709619 @default.
- W2779770929 cites W2341497066 @default.
- W2779770929 cites W2417750831 @default.
- W2779770929 cites W2460937040 @default.
- W2779770929 cites W2473640056 @default.
- W2779770929 cites W2535873859 @default.
- W2779770929 cites W2543927648 @default.
- W2779770929 cites W2556967412 @default.
- W2779770929 cites W2570685808 @default.
- W2779770929 cites W2579152745 @default.
- W2779770929 cites W2609920186 @default.
- W2779770929 cites W2738841453 @default.
- W2779770929 cites W2747648373 @default.
- W2779770929 cites W2785486371 @default.
- W2779770929 cites W2950557924 @default.
- W2779770929 cites W2950774971 @default.
- W2779770929 cites W2953106684 @default.
- W2779770929 cites W2963207607 @default.
- W2779770929 cites W2963307811 @default.
- W2779770929 cites W2963491874 @default.
- W2779770929 cites W2963920068 @default.
- W2779770929 cites W2963979376 @default.
- W2779770929 cites W2964095005 @default.
- W2779770929 cites W2964153729 @default.
- W2779770929 cites W3101998545 @default.
- W2779770929 cites W3106250896 @default.
- W2779770929 doi "https://doi.org/10.48550/arxiv.1712.08263" @default.
- W2779770929 hasPublicationYear "2017" @default.
- W2779770929 type Work @default.
- W2779770929 sameAs 2779770929 @default.
- W2779770929 citedByCount "0" @default.
- W2779770929 crossrefType "posted-content" @default.
- W2779770929 hasAuthorship W2779770929A5020806354 @default.
- W2779770929 hasAuthorship W2779770929A5058539660 @default.
- W2779770929 hasAuthorship W2779770929A5063048878 @default.
- W2779770929 hasAuthorship W2779770929A5089800102 @default.
- W2779770929 hasBestOaLocation W27797709291 @default.
- W2779770929 hasConcept C121332964 @default.
- W2779770929 hasConcept C154945302 @default.
- W2779770929 hasConcept C177918212 @default.
- W2779770929 hasConcept C2984842247 @default.
- W2779770929 hasConcept C31972630 @default.
- W2779770929 hasConcept C37736160 @default.
- W2779770929 hasConcept C41008148 @default.
- W2779770929 hasConcept C50644808 @default.
- W2779770929 hasConcept C62520636 @default.
- W2779770929 hasConcept C76155785 @default.
- W2779770929 hasConcept C94915269 @default.
- W2779770929 hasConceptScore W2779770929C121332964 @default.
- W2779770929 hasConceptScore W2779770929C154945302 @default.
- W2779770929 hasConceptScore W2779770929C177918212 @default.
- W2779770929 hasConceptScore W2779770929C2984842247 @default.
- W2779770929 hasConceptScore W2779770929C31972630 @default.
- W2779770929 hasConceptScore W2779770929C37736160 @default.
- W2779770929 hasConceptScore W2779770929C41008148 @default.
- W2779770929 hasConceptScore W2779770929C50644808 @default.
- W2779770929 hasConceptScore W2779770929C62520636 @default.
- W2779770929 hasConceptScore W2779770929C76155785 @default.
- W2779770929 hasConceptScore W2779770929C94915269 @default.
- W2779770929 hasLocation W27797709291 @default.
- W2779770929 hasOpenAccess W2779770929 @default.
- W2779770929 hasPrimaryLocation W27797709291 @default.
- W2779770929 hasRelatedWork W2516574342 @default.
- W2779770929 hasRelatedWork W3006111387 @default.
- W2779770929 hasRelatedWork W3016665419 @default.
- W2779770929 hasRelatedWork W3093347641 @default.
- W2779770929 hasRelatedWork W3113475396 @default.
- W2779770929 hasRelatedWork W3119467695 @default.
- W2779770929 hasRelatedWork W3194550388 @default.
- W2779770929 hasRelatedWork W4223932868 @default.
- W2779770929 hasRelatedWork W4225998895 @default.
- W2779770929 hasRelatedWork W4287867272 @default.
- W2779770929 isParatext "false" @default.
- W2779770929 isRetracted "false" @default.
- W2779770929 magId "2779770929" @default.
- W2779770929 workType "article" @default.