Controlling Over-generalization and its Effect on Adversarial Examples Generation and Detection

Convolutional Neural Networks (CNNs) significantly improve the state-of-the-art for many applications, especially in computer vision. However, CNNs still suffer from a tendency to confidently classify out-distribution samples from unknown classes into pre-defined known classes. Further, they are als...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	arXiv.org 2018-10
Hauptverfasser:	Abbasi, Mahdieh, Rajabi, Arezoo, Azadeh Sadat Mozafari, Bobba, Rakesh B, Gagne, Christian
Format:	Artikel
Sprache:	eng
Schlagworte:	Algorithms Artificial neural networks Computer vision Fitness Training
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!