Graph Representation Learning for Spatial Image Steganalysis

TL;DR

This paper proposes a novel graph-based neural network architecture for spatial image steganalysis, leveraging graph representations of image patches and attention mechanisms to improve detection performance.

Contribution

It introduces a new graph representation learning approach for steganalysis, translating images into graphs and applying attention networks, which is a novel application in this domain.

Findings

Achieves competitive performance with benchmark CNN models

Demonstrates the potential of graph learning in steganalysis

Highlights the effectiveness of attention mechanisms in feature discrimination

Abstract

In this paper, we introduce a graph representation learning architecture for spatial image steganalysis, which is motivated by the assumption that steganographic modifications unavoidably distort the statistical characteristics of the hidden graph features derived from cover images. In the detailed architecture, we translate each image to a graph, where nodes represent the patches of the image and edges indicate the local relationships between the patches. Each node is associated with a feature vector determined from the corresponding patch by a shallow convolutional neural network (CNN) structure. By feeding the graph to an attention network, the discriminative features can be learned for efficient steganalysis. Experiments indicate that the reported architecture achieves a competitive performance compared to the benchmark CNN model, which has shown the potential of graph learning for steganalysis.