Data Augmentation for JPEG Steganalysis
Authors:
- 7 December 2021
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Abstract
Deep Convolutional Neural Networks (CNNs) have performed remarkably well in JPEG steganalysis. However, they heavily rely on large datasets to avoid overfitting. Data augmentation is a popular technique to inflate the datasets available without collecting new images. For JPEG steganalysis, the augmentations predominantly used by researchers are limited to rotations and flips (D4 augmentations). This is due to the fact that the stego signal is erased by most augmentations used in computer vision. In this paper, we systematically survey a large number of other augmentation techniques and assess their benefit in JPEG steganalysis.Keywords
Funding Information
- NSF (2028119)
This publication has 12 references indexed in Scilit:
- BitMix: data augmentation for image steganalysisElectronics Letters, 2020
- Steganography by Minimizing Statistical Detectability: The cases of JPEG and Color Images2020
- Pixels-off: Data-augmentation Complementary Solution for Deep-learning Steganalysis2020
- Albumentations: Fast and Flexible Image AugmentationsInformation, 2020
- Deep learning in steganography and steganalysis2020
- Deep Learning Hierarchical Representations for Image SteganalysisIEEE Transactions on Information Forensics and Security, 2017
- Rich Models for Steganalysis of Digital ImagesIEEE Transactions on Information Forensics and Security, 2012
- Steganalysis of JPEG images using rich models2012
- Breaking HUGO – The Process Discovery2011
- Statistically undetectable jpeg steganography2007