A superior, open source suite of state-of-the-art steganalysis tools.

Techniques

Deepsteg performs visual attacks, structural attacks, and statistical attacks (including deep learning based attacks) to detect files hidden within images and other files. Eventually, we want to extract the hidden data from these files.

1

RS Analysis

Detects randomly scattered LSB embedding in grayscale and colour images by inspecting the differences in the number of regular and singular groups for the LSB and ’shifted’ LSB plane.

2

Chi-square attack

Statistical analysis of pairs of values (PoV’s) exchanged during LSB embedding.

3

Sample Pair Analysis

Based on a finite state machine whose states are selected multisets of sample pairs called trace multisets.

4

Difference histogram analysis

Statistical attack on an image’s histogram, measuring the correlation between the least significant and all other bit planes.

5

Primary Sets

Based on a statistical identity related to certain sets of pixels in an image.

6

Deep learning based methods

Building a neural network that is trained to detect steganography.

Fighting crime online

Steganography is used by criminals to send messages to other criminals secretly. It is used by terrorists, pedophiles, and other criminals. Spreading awareness and providing tools to expose hidden data can help fight crime.

We are still under construction.

Currently, we are still writing the software for our MVP. Stay tuned for updates.