.png){kind=icon}
Return
KeyMorph: Learning parameter-efficient network activation for deep steganography
.png){kind=icon}
PRE
AI
DOI：10.1016/j.patcog.2026.115155.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• KeyMorph enables steganography without transmitting large or suspicious parameters.
• Achieving stronger structural and functional indistinguishability from standard CNNs.
• Designing KD-Conv to enable seamless network morphing.
• Superior visual quality and undetectability on DIV2K, COCO and ImageNet datasets.
Keywords:
Deep learning
Steganography
Key-driven morphing
Transmission of steganographic tools
Journal
IF:
7.6
Papers:
1.3W
Citations:
4.5W
Organization
No organization information available
Cited Papers
No cited papers available