Чланци са овлашћеним јавним приступом - Andras RozsaСазнајте више
Доступно негде: 9
Adversarial diversity and hard positive generation
A Rozsa, EM Rudd, TE Boult
Proceedings of the IEEE Conference on Computer Vision and Pattern …, 2016
Овлашћења: US National Science Foundation
A Survey of Stealth Malware Attacks, Mitigation Measures, and Steps Toward Autonomous Open World Solutions
EM Rudd, A Rozsa, M Günther, TE Boult
IEEE Communications Surveys & Tutorials 19 (2), 1145-1172, 2017
Овлашћења: US National Science Foundation
Assessing threat of adversarial examples on deep neural networks
A Graese, A Rozsa, TE Boult
2016 15th IEEE International Conference on Machine Learning and Applications …, 2016
Овлашћења: US National Science Foundation
Are accuracy and robustness correlated
A Rozsa, M Günther, TE Boult
2016 15th IEEE International Conference on Machine Learning and Applications …, 2016
Овлашћења: US National Science Foundation
AFFACT: Alignment-free facial attribute classification technique
M Günther, A Rozsa, TE Boult
2017 IEEE International Joint Conference on Biometrics (IJCB), 90-99, 2017
Овлашћења: US National Science Foundation, US Department of Defense, US Office of the …
Facial attributes: Accuracy and adversarial robustness
A Rozsa, M Günther, EM Rudd, TE Boult
Pattern Recognition Letters 124, 100-108, 2019
Овлашћења: US National Science Foundation, US Department of Defense, US Office of the …
Are facial attributes adversarially robust?
A Rozsa, M Günther, EM Rudd, TE Boult
2016 23rd International Conference on Pattern Recognition (ICPR), 3121-3127, 2016
Овлашћења: US National Science Foundation
LOTS about attacking deep features
A Rozsa, M Günther, TE Boult
2017 IEEE International Joint Conference on Biometrics (IJCB), 168-176, 2017
Овлашћења: US National Science Foundation, US Department of Defense, US Office of the …
On the Robustness of Deep Neural Networks
M Günther, A Rozsa, TE Boult
Овлашћења: US National Science Foundation, US Department of Defense, US Office of the …
Публикације и информације о финансирању аутоматски одређује рачунарски програм