Deutsch
 
Hilfe Datenschutzhinweis Impressum
  DetailsucheBrowse

Datensatz

DATENSATZ AKTIONENEXPORT
  Accurate, reliable and fast robustness evaluation

Brendel, W., Rauber, J., Kümmerer, M., Ustyuzhaninov, I., & Bethge, M. (2019). Accurate, reliable and fast robustness evaluation. In H. Wallach, H. Larochelle, A. Beygelzimer, F. d'Alché-Buc, E. Fox, & R. Garnett (Eds.), Advances in Neural Information Processing Systems 32 (pp. 12841-12851).

Item is

Basisdaten

einblenden: ausblenden:
Genre: Konferenzbeitrag

Dateien

einblenden: Dateien
ausblenden: Dateien
:
NEURIPS-2019-Brendel.pdf (Zusammenfassung), 16MB
Name:
NEURIPS-2019-Brendel.pdf
Beschreibung:
-
OA-Status:
Sichtbarkeit:
Öffentlich
MIME-Typ / Prüfsumme:
application/pdf / [MD5]
Technische Metadaten:
Copyright Datum:
-
Copyright Info:
-
Lizenz:
-

Externe Referenzen

einblenden:
ausblenden:
Beschreibung:
-
OA-Status:

Urheber

einblenden:
ausblenden:
 Urheber:
Brendel, W, Autor
Rauber, J, Autor
Kümmerer, M, Autor
Ustyuzhaninov, I, Autor
Bethge, M1, 2, Autor           
Affiliations:
1Research Group Computational Vision and Neuroscience, Max Planck Institute for Biological Cybernetics, Max Planck Society, ou_1497805              
2Max Planck Institute for Biological Cybernetics, Max Planck Society, Spemannstrasse 38, 72076 Tübingen, DE, ou_1497794              

Inhalt

einblenden:
ausblenden:
Schlagwörter: -
 Zusammenfassung: Throughout the past five years, the susceptibility of neural networks to minimal adversarial perturbations has moved from a peculiar phenomenon to a core issue in Deep Learning. Despite much attention, however, progress towards more robust models is significantly impaired by the difficulty of evaluating the robustness of neural network models. Today's methods are either fast but brittle (gradient-based attacks), or they are fairly reliable but slow (score- and decision-based attacks). We here develop a new set of gradient-based adversarial attacks which (a) are more reliable in the face of gradient-masking than other gradient-based attacks, (b) perform better and are more query efficient than current state-of-the-art gradient-based attacks, (c) can be flexibly adapted to a wide range of adversarial criteria and (d) require virtually no hyperparameter tuning. These findings are carefully validated across a diverse set of six different models and hold for L2 and L-infinity in both targeted as well as untargeted scenarios. Implementations will be available in all major toolboxes (Foolbox, CleverHans and ART). We hope that this class of attacks will make robustness evaluations easier and more reliable, thus contributing to more signal in the search for more robust machine learning models.

Details

einblenden:
ausblenden:
Sprache(n):
 Datum: 2019-072019-092019-12
 Publikationsstatus: Online veröffentlicht
 Seiten: -
 Ort, Verlag, Ausgabe: -
 Inhaltsverzeichnis: -
 Art der Begutachtung: -
 Identifikatoren: -
 Art des Abschluß: -

Veranstaltung

einblenden:
ausblenden:
Titel: Thirty-third Conference on Neural Information Processing Systems (NeurIPS 2019)
Veranstaltungsort: Vancouver, Canada
Start-/Enddatum: 2019-12-09 - 2019-12-13

Entscheidung

einblenden:

Projektinformation

einblenden:

Quelle 1

einblenden:
ausblenden:
Titel: Advances in Neural Information Processing Systems 32
Genre der Quelle: Konferenzband
 Urheber:
Wallach, H, Herausgeber
Larochelle, H, Herausgeber
Beygelzimer , A, Herausgeber
d'Alché-Buc, F, Herausgeber
Fox, E, Herausgeber
Garnett, R, Herausgeber
Affiliations:
-
Ort, Verlag, Ausgabe: -
Seiten: - Band / Heft: - Artikelnummer: - Start- / Endseite: 12841 - 12851 Identifikator: -