Transcription of Adversarial Generative Nets: Neural Network …
{{id}} {{{paragraph}}}
Adversarial Generative nets : Neural NetworkAttacks on State-of-the-Art Face RecognitionMahmood Sharif, Sruti Bhagavatula, Lujo BauerCarnegie Mellon University{mahmoods, srutib, K. ReiterUniversity of North Carolina at Chapel In this paper we show that misclassification attacksagainst face-recognition systems based on deep Neural networks(DNNs) are more dangerous than previously demonstrated, evenin contexts where the adversary can manipulate only her physicalappearance (versus directly manipulating the image input to theDNN). Specifically, we show how to create eyeglasses that, whenworn, can succeed in targeted (impersonation) or untargeted(dodging) attacks while improving on previous work in one ormore of three facets: (i) inconspicuousness to onlooking observers,which we test through a user stu}
Adversarial Generative Nets: Neural Network Attacks on State-of-the-Art Face Recognition Mahmood Sharif, Sruti Bhagavatula, Lujo Bauer Carnegie Mellon University
Domain:
Source:
Link to this page:
Please notify us if you found a problem with this document:
{{id}} {{{paragraph}}}