Adversarial Attack on Pre-Trained CNN Model

Description

Course: Machine Learning (Fall 2022)
Task: Create an adversarial example for each of the 10 test images and use Pre-Trained models to classify the 10 test images.

Pre-trained model that was used:

• InceptionV31
• DenseNet121
• MobileNetV2

Adversarial attack that was used:

• Fast Gradient Sign Method (FGSM)