Reversible data hiding scheme using prediction neural network and adaptive modulation mapping

Multimedia Tools and Applications(2024)

Cited 0|Views4
No score
Abstract
Prediction error expansion (PEE) is an attractive approach for reversible data hiding (RDH). The key issue for PEE-based RDH is to improve the prediction accuracy and design a better modulation mapping. This paper proposes a novel prediction error modulation (PEM) scheme, which comprises a prediction neural network and a modulation mapping rule generation algorithm. We use the multi-scale feature extraction (MSFE) module and the residual dense networks (RDN) to construct the prediction neural network. In the proposed RDH scheme, the cover image is divided into two parts, the reference pixel set and the cover pixel set, based on the chessboard pattern. The prediction network serves to predict the values of the cover pixel set using the reference pixel set. An optimization model is set up to adaptively determine the optimal modulation mapping for data hiding based on the specific prediction-error histogram and payload constraint. Experimental results show the superiority of the proposed prediction neural network and the optimized modulation mapping compared with related works.
More
Translated text
Key words
Reversible data hiding,Prediction error modulation,Neural networks,Adaptive modulation mapping
AI Read Science
Must-Reading Tree
Example
Generate MRT to find the research sequence of this paper
Chat Paper
Summary is being generated by the instructions you defined