Adversarial Extra in ZOO.md

title	date	comments	author	categories	tags
Adversarial Extra	2020-05-27 16:20:00 -0700	true	Yi-Wei	meeting read note	Adversarial note

tags: adversarial

Adversarial

Adversarial example in some norm

在之前的ZOO可以看到我們創造的是$norm_2$的adversarial example 在JSMA創造的是$norm_\infty$ 在EDA創造的是$norm_1$的例子，

Adversarial examples in more type

• The use in CNN+RNN :

• The use in speech recognition

• The use in NLP with Sentiment Analysis,Fake-News Detection, Spam Filtering.

The improve of the ZOO -> AUTOZOO

• (Auto encoder based Z eroth Order Optimization Method for Attacking Black box Neural Networks) 主要做的有兩點：

1. 使用了更新的gradient算法: The Old one The New one : $b \in (1, dimensions)$ The Sucessful rate in iterations

2. Use the dimension reduction: Here is the two method Left : Autoencoder Right: Biliner

STRUCTURED ADVERSARIAL ATTACK

ADMM used to solve the subproblem ‘ostrich’ is the original label, and ‘unicycle’ is the misclassified label.

將圖片分成一些區塊，並且將那些區塊做loss處理，並且用ADMM將方程式做處理，處理過後的方程式就不列上了，有點長 連結在此