© The Institution of Engineering and Technology
Convolutional neural networks for image steganalysis demonstrate better performances with employing concepts from high-level vision tasks. The major employed concept is to use data augmentation to avoid overfitting due to limited data. To augment data without damaging the message embedding, only rotating multiples of or horizontally flipping are used in steganalysis, which generates eight fixed results from one sample. To overcome this limitation, the authors propose BitMix, a data augmentation method for spatial image steganalysis. BitMix mixes a cover and stego image pair by swapping the random patch and generates an embedding adaptive label with the ratio of the number of pixels modified in the swapped patch to those in the cover–stego pair. The authors explore optimal hyperparameters, the ratio of applying BitMix in the mini-batch, and the size of the bounding box for swapping patch. The results reveal that using BitMix improves the performance of spatial image steganalysis and better than other data augmentation methods.
References
-
-
1)
-
5. Ye, J., Ni, J., Yi, Y.: ‘Deep learning hierarchical representations for image steganalysis’, IEEE Trans. Inf. Forensics Sec., 2017, 12, (11), pp. 2545–2557 (doi: 10.1109/TIFS.2017.2710946).
-
2)
-
3. Boroumand, M., Chen, M., Fridrich, J.: ‘Deep residual network for steganalysis of digital images’, IEEE Trans. Inf. Forensics Sec., 2019, 14, (5), pp. 1181–1193 (doi: 10.1109/TIFS.2018.2871749).
-
3)
-
4. Zhang, R., Zhu, F., Liu, J., et al: ‘Depth-wise separable convolutions and multi-level pooling for an efficient spatial CNN-based steganalysis’, IEEE Trans. Inf. Forensics Sec., 2020, 15, pp. 1138–1150 (doi: 10.1109/TIFS.2019.2936913).
-
4)
-
6. Yedroudj, M., Comby, F., Chaumont, M.: ‘Yedroudj-net: an efficient CNN for spatial steganalysis’. 2018 IEEE Int. Conf. on Acoustics Speech and Signal Processing (ICASSP), Calgary, AB, Canada, 2018, pp. 2092–2096.
-
5)
-
9. Yoo, J., Ahn, N., Sohn, K.A.: ‘, 2020.
-
6)
-
8. Yun, S., Han, D., Oh, S.J., et al: ‘Cutmix: regularization strategy to train strong classifiers with localizable features’. Proc. IEEE Int. Conf. on Computer Vision, Seoul, Republic of Korea, 2019, pp. 6023–6032.
-
7)
-
2. Xu, G., Wu, H.Z., Shi, Y.Q.: ‘Ensemble of CNNs for steganalysis: an empirical study’. Proc. Fourth ACM Workshop on Information Hiding and Multimedia Security, Vigo, Galicia, Spain, 2016, pp. 103–107.
-
8)
-
1. Xu, G., Wu, H.Z., Shi, Y.Q.: ‘Structural design of convolutional neural networks for steganalysis’, IEEE Signal Process. Lett., 2016, 23, (5), pp. 708–712 (doi: 10.1109/LSP.2016.2548421).
-
9)
-
7. Zhang, H., Cisse, M., Dauphin, Y.N., et al: ‘, 2017.
http://iet.metastore.ingenta.com/content/journals/10.1049/el.2020.1951
Related content
content/journals/10.1049/el.2020.1951
pub_keyword,iet_inspecKeyword,pub_concept
6
6