Downloads
Steganography
- Steganographic Algorithms
- Syndrome-Trellis Codes Toolbox
- Steganography Design
- Gibbs Construction in Steganography
- Simulator of nsF5
- Perturbed Quantization
Steganalysis
- Feature Extractors
- Are we there yet?
- Ensemble Classifier
- Low-complexiy Linear Classifier
- LRT Linear Classifier
- Histogram Layer
- Explicit Feature Maps
- SPAM features
- Extractor of 274/548 Merged Features
- Structural LSB Detectors
- Quantitative Steganalysis Using Rich Models
Digital Forensics
Image Database
Histogram Layer, Moving Convolutional Neural Networks Towards Feature-based Steganalysis
Description
On this page, you can download a C++ implementation of the histogram layer as described in [1], using Caffe [2]. Caffe is a deep learning framework made with expression, speed, and modularity in mind. It is developed by the Berkeley Vision and Learning Center (BVLC) and by community contributors.
Disclaimer
This program is available for public use. Please, remember to recognize our work by citing [1].
Thank you.
Paper abstract
Feature-based steganalysis has been an integral tool for detecting the presence of steganography in communication channels for a long time. In this paper, we explore the possibility to utilize powerful optimization algorithms available in convolutional neural network packages to optimize the design of rich features. To this end, we implemented a new layer that simulates the formation of histograms from truncated and quantized noise residuals computed by convolution. Our goal is to show the potential to compactify and further optimize existing features, such as the projection spatial rich model (PSRM).
Contact
- Vahid Sedighi - vsedigh1 (at) binghamton (dot) edu
- Jessica Fridrich - fridrich (at) binghamton (dot) edu
Download
C++/CUDA implementation of the histogram layer. Follow the ReadMe.txt for proper embedding of the layer within Caffe.
- - download: [ histogram_layer.zip ] (5.28KB)
Sample CNN model prototxt file for modeling PSRM sub-models as described in [1]. Please note that you should specify the path to your training database in the model. Also, you need to embed a proper kernel (KV) as the first convolutional layer ("conv0"), upon instantiating and before starting the training procedure of the model.
- - download: [ hist.prototxt.zip ] (1KB)
References
[1] V. Sedighi, and J. Fridrich, Histogram Layer, Moving Convolutional Neural Networks Towards Feature-based Steganalysis. Proc. IS&T, Electronic Imaging, Media Watermarking, Security, and Forensics 2017, San Francisco, CA, January 29–February 2, 2017.
[2] Y. Jia, E. Shelhamer, J. Donahue, S. Karayev, J. Long, R. B. Girshick, S. Guadarrama, and T. Darrell, Caffe: Convolutional architecture for fast feature embedding. CoRR, abs/1408.5093, 2014.
Last update: January 2017