Enhancing image steganalysis via multi-source domain invariant feature learning
The Visual Computer · 2026-06 · 42 (8)
DOI: 10.1007/s00371-026-04537-4 ↗Kurz gefasst
Fachartikel in der Zeitschrift The Visual Computer, online erschienen am 08.06.2026. Verfasst von Xueyi Ye und 4 weiteren Autor:innen. Die Arbeit nennt 62 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: National Natural Science Foundation of China.
Abstract (Originalsprache)
Image steganalysis faces significant challenges due to cover source mismatch (CSM) caused by diverse steganographic algorithms. This paper proposes a model that learns both intra-domain- and inter-domain-invariant features across multiple source domains to improve generalization. Distillation learning extracts rich spatial and frequency domain features, while second-order statistics alignment minimizes feature discrepancies. Dynamic weight assignment balances optimization objectives, enhancing model performance. Experimental results demonstrate up to a 4.87% improvement in detection accuracy under mismatched scenarios, showcasing the model’s practical applicability. The relate code and datasets will be posted at https://github.com/ailiaoye/msd.git.
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Ye, X., Tan, R., Chen, H., Wang, H., & Chen, H. (2026). Enhancing image steganalysis via multi-source domain invariant feature learning. The Visual Computer, 42(8). https://doi.org/10.1007/s00371-026-04537-4
Vancouver
Ye X, Tan R, Chen H, Wang H, Chen H. Enhancing image steganalysis via multi-source domain invariant feature learning. The Visual Computer. 2026;42(8). doi:10.1007/s00371-026-04537-4
BibTeX
@article{Ye2026,
author = {Ye, Xueyi and Tan, Ruijie and Chen, Haiying and Wang, Hao and Chen, Huahua},
title = {Enhancing image steganalysis via multi-source domain invariant feature learning},
journal = {The Visual Computer},
year = {2026},
volume = {42},
number = {8},
doi = {10.1007/s00371-026-04537-4}
}
EUVIMED