Which ONE of the following statements is a CORRECT adversarial example in the context of machine learning systems that are working on image classifiers?

The correct answer is D. These examples are model specific and are not likely to cause another model trained on same. Adversarial examples are often model-specific, meaning that they exploit the specific weaknesses of a particular model. While some adversarial examples might transfer between models, many are tailored to the specific model they were generated for and may not affect other models…

Testing, Debugging, and Deployment

Question

Options

ABlack box attacks based on adversarial examples create an exact duplicate model of the original.
BThese attack examples cause a model to predict the correct class with slightly less accuracy even
CThese attacks can't be prevented by retraining the model with these examples augmented to the
DThese examples are model specific and are not likely to cause another model trained on same

How the community answered

(50 responses)

A
10% (5)
B
6% (3)
C
2% (1)
D
82% (41)

Explanation

Adversarial examples are often model-specific, meaning that they exploit the specific weaknesses of a particular model. While some adversarial examples might transfer between models, many are tailored to the specific model they were generated for and may not affect other models trained on the same task.

Topics

#adversarial examples#black box attacks#ML security#image classification

Community Discussion

No community discussion yet for this question.

Full CT-AI Practice