Skip to content
AI-Driven Image Encryption
Feb 2025 - May 2025

AI-Driven Image Encryption

Novel GAN-LSTM encryption for one-pass workflow

01. The Challenge

Traditional image encryption methods are computationally expensive and vulnerable to attacks.

02. The Solution

Developed novel GAN-LSTM based encryption system for secure, efficient one-pass image encryption.


Overview

Standard encryption like AES is secure but not optimized for the massive size of image data. This project explores using Generative Adversarial Networks (GANs) combined with LSTMs to create a chaotic encryption map.


The Approach

The GAN generates a chaotic key map based on a seed. The LSTM then processes the image pixels in sequence, using the chaotic map to permute and diffuse the pixel values.


Performance

The one-pass workflow significantly reduces encryption time compared to traditional multi-round block ciphers, making it suitable for real-time video streams.


System Architecture

GAN for encryption key generation
LSTM for temporal dependencies
PyTorch implementation
Docker containerization

Project Links

Technologies

PyTorchGANLSTMDockerPython

Key Impact

One-pass encryption workflow
Enhanced security
Reduced computational overhead