On the claim for the existence of "Adversarial examples" in deep learning neural networks
Ημερομηνία
2014ISBN
978-989-758-054-3Εκδότης
INSTICC PressSource
NCTA 2014 - Proceedings of the International Conference on Neural Computation Theory and Applications6th International Conference on Neural Computation Theory and Applications, NCTA 2014, Part of the 6th International Joint Conference on Computational Intelligence, IJCCI 2014
Pages
306-309Google Scholar check
Keyword(s):
Metadata
Εμφάνιση πλήρους εγγραφήςΕπιτομή
A recent article in which it is claimed that adversarial examples exist in deep artificial neural networks (ANN) is critically examined. The newly discovered properties of ANNs are critically evaluated. Specifically, we point that adversarial examples can be serious problems in critical applications of pattern recognition. Also, they may stall the further development of artificial neural networks. We challenge the absolute existence of these examples, as this has not been universally proven yet. We also suggest that ANN structures, that correctly recognize adversarial examples, can be developed.