DHU-Net: High-capacity binary data hiding network based on improved U-Net
- Authors
- Duan, Xintao; Li, Chun; Wei, Bingxin; Wu, Guoming; Qin, Chuan; Nam, Haewoon
- Issue Date
- Apr-2024
- Publisher
- Elsevier B.V.
- Keywords
- Binary data hiding; DHU-Net; Flow encoding; Losslessly
- Citation
- Neurocomputing, v.576, pp 1 - 16
- Pages
- 16
- Indexed
- SCIE
SCOPUS
- Journal Title
- Neurocomputing
- Volume
- 576
- Start Page
- 1
- End Page
- 16
- URI
- https://scholarworks.bwise.kr/erica/handle/2021.sw.erica/118317
- DOI
- 10.1016/j.neucom.2024.127314
- ISSN
- 0925-2312
1872-8286
- Abstract
- Due to limited data-hiding capacity and low extraction accuracy, most existing data hiding schemes have difficulty in high capacity hiding and lossless extraction of binary data. This paper proposes a novel binary data hiding network (DHU-Net) based on improved U-Net and flow encoding, which deepens the network structure and facilitates lossless extraction of high-capacity binary data. The BN layer is added to address overfitting and prevent vanishing gradient or exploding gradient. Furthermore, a novel binary-to-spatial domain mapping method is proposed to map the binary data to the spatial domain. Experimental results show that our scheme can achieve lossless extraction of binary data while maintaining low distortion and sufficient security. DHU-Net provides a promising solution for secure and efficient transmission of binary data through data hiding. © 2024 Elsevier B.V.
- Files in This Item
-
- Appears in
Collections - COLLEGE OF ENGINEERING SCIENCES > SCHOOL OF ELECTRICAL ENGINEERING > 1. Journal Articles
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.