This project is the implementation of research paper “LEARNING TO PROTECT COMMUNICATIONS WITH ADVERSARIAL NEURAL CRYPTOGRAPHY”
It is the Computer Networks Project which covers the concepts of Network security, Encryption, Decryption, Convolutional Neural Networks and Tensorflow.
-
We make neural networks to learn to use secret keys to protect information from several other neural networks. Specifically, we focus on ensuring confidentiality properties in a multiagent system, and we specify those properties in terms of an adversary.
-
Thus, a system may consist of neural networks named Alice and Bob, and we aim to limit what a third neural network named Eve learns from eaves dropping on the communication between Alice and Bob.
-
We do not prescribe specific cryptographic algorithms to these neural networks; instead, we train end-to-end, adversarially.
-
We demonstrate that the neural networks can learn how to perform forms of encryption and decryption, and also how to apply these operations selectively in order to meet confidentiality goals.
- Python 3
- Tensorflow