Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

FIIH: Fully Invertible Image Hiding for Secure and Robust

Lang Huang, Lin Huo, Zheng Gan, Xinrong He

TL;DR

FIIH addresses the non-invertible nature of prior image hiding methods by introducing a fully invertible framework that hides secret images using a shared invertible neural network, preserving information in both the hidden and revealed stages. The approach decomposes the secret via $DWT$, embeds it into the high-frequency domain through an INN with shared parameters, and recovers the secret using an inverse path; robustness is further strengthened with a Field Pixel Filling-based dropout strategy and a noise/JPEG augmentation protocol. Key contributions include achieving fully invertible data hiding, improving resistance to deep-learning steganalysis, and delivering state-of-the-art performance in fidelity (PSNR/SSIM), robustness against Gaussian noise and JPEG, and security against SRNet-based steganalysis. The results demonstrate strong practical impact for covert transmission and secure image hiding in challenging, lossy networks, with future work aiming at multi-image invertible hiding and broader domain applications.

Abstract

Image hiding is the study of techniques for covert storage and transmission, which embeds a secret image into a container image and generates stego image to make it similar in appearance to a normal image. However, existing image hiding methods have a serious problem that the hiding and revealing process cannot be fully invertible, which results in the revealing network not being able to recover the secret image losslessly, which makes it impossible to simultaneously achieve high fidelity and secure transmission of the secret image in an insecure network environment. To solve this problem,this paper proposes a fully invertible image hiding architecture based on invertible neural network,aiming to realize invertible hiding of secret images,which is invertible on both data and network. Based on this ingenious architecture, the method can withstand deep learning based image steganalysis. In addition, we propose a new method for enhancing the robustness of stego images after interference during transmission. Experiments demonstrate that the FIIH proposed in this paper significantly outperforms other state-of-the-art image hiding methods in hiding a single image, and also significantly outperforms other state-of-the-art methods in robustness and security.

FIIH: Fully Invertible Image Hiding for Secure and Robust

TL;DR

FIIH addresses the non-invertible nature of prior image hiding methods by introducing a fully invertible framework that hides secret images using a shared invertible neural network, preserving information in both the hidden and revealed stages. The approach decomposes the secret via , embeds it into the high-frequency domain through an INN with shared parameters, and recovers the secret using an inverse path; robustness is further strengthened with a Field Pixel Filling-based dropout strategy and a noise/JPEG augmentation protocol. Key contributions include achieving fully invertible data hiding, improving resistance to deep-learning steganalysis, and delivering state-of-the-art performance in fidelity (PSNR/SSIM), robustness against Gaussian noise and JPEG, and security against SRNet-based steganalysis. The results demonstrate strong practical impact for covert transmission and secure image hiding in challenging, lossy networks, with future work aiming at multi-image invertible hiding and broader domain applications.

Abstract

Image hiding is the study of techniques for covert storage and transmission, which embeds a secret image into a container image and generates stego image to make it similar in appearance to a normal image. However, existing image hiding methods have a serious problem that the hiding and revealing process cannot be fully invertible, which results in the revealing network not being able to recover the secret image losslessly, which makes it impossible to simultaneously achieve high fidelity and secure transmission of the secret image in an insecure network environment. To solve this problem,this paper proposes a fully invertible image hiding architecture based on invertible neural network,aiming to realize invertible hiding of secret images,which is invertible on both data and network. Based on this ingenious architecture, the method can withstand deep learning based image steganalysis. In addition, we propose a new method for enhancing the robustness of stego images after interference during transmission. Experiments demonstrate that the FIIH proposed in this paper significantly outperforms other state-of-the-art image hiding methods in hiding a single image, and also significantly outperforms other state-of-the-art methods in robustness and security.
Paper Structure (28 sections, 19 equations, 10 figures, 6 tables)

This paper contains 28 sections, 19 equations, 10 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Robust Image Hiding
  4. Invertible Neural Network
  5. Method
  6. Network architecture
  7. Invertible neural network architecture
  8. Robustness Module
  9. Gaussian noise
  10. Dropout
  11. Jpeg compression
  12. Loss function
  13. Experiment
  14. Experimental Settings
  15. Data setting
  16. ...and 13 more sections

Figures (10)

  • Figure 1: illustrates the difference between our image hiding method and traditional invertible neural network methods.
  • Figure 2: Architecture of FIIH's overall model.
  • Figure 3: Architecture of invertible neural networks.
  • Figure 4: The process of running four-field pixel filling.
  • Figure 5: Visual comparison of Hinet, FMIN and our method on single image hiding, the first row of the right part is the cover image and the secret image from left to right, the second row is the loaded image and the secret image extracted from the loaded image, and the third row is the residual image *20.
  • ...and 5 more figures