detection-of-model-inversion-attack

The <L2_norm> fold contains graphs of l2-norm value for number 0-9 with their first 300 iterations.

The <cosine_similarity> fold contains graphs of cosine similarity value for number 0-9 with their first 300 iterations.

The <MIFace_MNIST(CIFAR10).ipynb> is the file that applies MI attack on MNIST(CIFAR10) dataset.

The <myvgg11_cifar10.ipynb>vis the file that trains a vgg11 network to classify CIFAR10 dataset.

The <detector.ipynb> is using combined method to detect MI attack and save record to .csv file.

Name		Name	Last commit message	Last commit date
Latest commit History 18 Commits
L2 norm		L2 norm
cosine_similarity		cosine_similarity
LICENSE		LICENSE
MIFace_CIFAR10.ipynb		MIFace_CIFAR10.ipynb
MIFace_MNIST.ipynb		MIFace_MNIST.ipynb
README.md		README.md
detector.ipynb		detector.ipynb
main.md		main.md
mi_attack.ipynb		mi_attack.ipynb
myvgg11_cifar10.ipynb		myvgg11_cifar10.ipynb

Provide feedback