BriefGPT.xyz
Nov, 2017
LatentPoison - 针对潜在空间的对抗性攻击
LatentPoison - Adversarial Attacks On The Latent Space
HTML
PDF
Antonia Creswell, Anil A. Bharath, Biswa Sengupta
TL;DR
研究深度变分自编码器的潜在空间的稳健性，证明可以扰动潜在空间并使分类概率近乎相等，从而保证解码器输出不受攻击影响。
Abstract
robustness
and
security
of
machine learning
(ML) systems are intertwined, wherein a non-robust ML system (classifiers, regressors, etc.) c
→