RT Conference Proceedings
T1 Encoding generative adversarial networks for defense against image classification attacks
A1 Rodríguez Rodríguez, José Antonio
A1 Pérez Bravo, José María
A1 García-González, Jorge
A1 Molina-Cabello, Miguel Ángel
A1 Thurnhofer-Hemsi, Karl
A1 López-Rubio, Ezequiel
K1 Inteligencia artificial
K1 Algoritmos
K1 Redes neuronales (Informática)
K1 Aprendizaje automático (Inteligencia artificial)
AB Image classification has undergone a revolution in recent years due to the high performance of new deep learning models. However, severe security issues may impact the performance of these systems. In particular, adversarial attacks are based on modifying input images in a way that is imperceptible for human vision, so that deep learning image classifiers are deceived. This work proposes a new deep neural networkmodel composed of an encoder and a Generative Adversarial Network (GAN). The former encodes a possibly malformed input image into a latent vector, while the latter generates a reconstructed image from thelatent vector. Then the reconstructed image can be reliably classified because our model removes the deleterious effects of the attack. The experiments carried out were designed to test the proposed approachagainst the Fast Gradient Signed Method attack. The obtained results demonstrate the suitability of our approach in terms of an excellent balance between classification accuracy and computational cost.
YR 2022
FD 2022
LK https://hdl.handle.net/10630/24396
UL https://hdl.handle.net/10630/24396
LA eng
NO Universidad de Málaga. Campus de Excelencia Internacional Andalucía Tech.
DS RIUMA. Repositorio Institucional de la Universidad de Málaga
RD 20 ene 2026