This was my End Semester Project for Computer Vision. In this project, we had to fool the CNN (in our case AlexNet) to predict another label for image by minimally updating it i.e. updating the image in such a way that a human eye can not differentiate between the two images.
This is a python3 based code that has been tested on python 3.5. Additionally it requires following libraries: -
- Numpy
- OpenCV
- PyTorch == 0.3
- Matplotlib
More over, to run this project, you will need AlexNet weights. You can download them from this link and place them in the models directory.
You can find an explaination of this code on my blog post here.