LATIDIA · Ciberseguridad

RPP: un marco de detección de muestra envenenada certificado para ataques de puerta trasera bajo desequilibrio de conjunto de datos

arXiv: 2602.00183v2Tipo de anuncio: reemplazar Resumen: Las redes neuronales profundas son altamente susceptibles a los ataques de puerta trasera, sin embargo, la mayoría de los métodos de defensa hasta la fecha dependen de datos equilibrados, pasando por alto el desequilibrio de clases generalizado en re

La noticia

arXiv:2602.00183v2 Announce Type: replace Abstract: Deep neural networks are highly susceptible to backdoor attacks, yet most defense methods to date rely on balanced data, overlooking the pervasive class imbalance in real-world scenarios that can amplify backdoor threats. This paper presents the first in-depth investigation of how the dataset imbalance amplifies backdoor vulnerability, showing that (i) the imbalance induces a majority-class bias that increases susceptibility and (ii) conventional defenses degrade significantly as the imbalance grows. To address this, we propose Randomized Probability Perturbation (RPP), a certified poisoned-sample detection framework that operates in a black-box setting using only model output probabilities. For any inspected sample, RPP determines whether the input has been