DeepMIH: Deep Invertible Network for Multiple Image Hiding

Guan, Zhenyu; Jing, Junpeng; Deng, Xin; Xu, Mai; Jiang, Lai; Zhang, Zhou; Li, Yipeng

doi:10.1109/tpami.2022.3141725

Cited by 111 publications

(56 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Existing data hiding schemes of high embedding rates hide full-size images within images using DNNs, which can realize more than 24 bpp. According to the different model structures, we can classify these schemes into three categories: encoder-decoder structure [16], [51], [52], [19], [17], [53], generative adversarial network (GAN) structure [54], [55], [18], and invertible neural network structure (INN) [20], [56].…”

Section: B Data Hiding Of High Embedding Ratesmentioning

confidence: 99%

“…Lu et al [20] used a INN to play both the hiding and revealing networks simultaneously and realized multipleimage hiding by concatenating multiple secret images in the channel dimension. Guan et al [56] developed an invertible hiding neural network (IHNN) that can be cascaded multiple times. When hiding multiple secret images, IHNN is able to guide the current image hiding based on the previous image hiding results.…”

Section: B Data Hiding Of High Embedding Ratesmentioning

confidence: 99%

“…Dataset. We use ImageNet and COCO to evaluate all involved hiding schemes [16], [17], [18], [20], [56]. For each dataset, we randomly select 10,000 images, including 8,000 Implementation of LDH and Baselines.…”

Section: A Configurationsmentioning

confidence: 99%

See 2 more Smart Citations

Smaller Is Bigger: Rethinking the Embedding Rate of Deep Hiding

Li¹,

Liu²,

Guo³

et al. 2023

Preprint

View full text Add to dashboard Cite

Deep hiding, concealing secret information using Deep Neural Networks (DNNs), can significantly increase the embedding rate and improve the efficiency of secret sharing. Existing works mainly force on designing DNNs with higher embedding rates or fancy functionalities. In this paper, we want to answer some fundamental questions: how to increase and what determines the embedding rate of deep hiding. To this end, we first propose a novel Local Deep Hiding (LDH) scheme that significantly increases the embedding rate by hiding large secret images into small local regions of cover images. Our scheme consists of three DNNs: hiding, locating, and revealing. We use the hiding network to convert a secret image in a small imperceptible compact secret code that is embedded into a random local region of a cover image. The locating network assists the revealing process by identifying the position of secret codes in the stego image, while the revealing network recovers all full-size secret images from these identified local regions. Our LDH achieves an extremely high embedding rate, i.e., 16×24 bpp and exhibits superior robustness to common image distortions. We also conduct comprehensive experiments to evaluate our scheme under various system settings. We further quantitatively analyze the trade-off between the embedding rate and image quality with different image restoration algorithms.

show abstract

Section: B Data Hiding Of High Embedding Ratesmentioning

confidence: 99%

Section: B Data Hiding Of High Embedding Ratesmentioning

confidence: 99%

See 1 more Smart Citation

Smaller Is Bigger: Rethinking the Embedding Rate of Deep Hiding

Li¹,

Liu²,

Guo³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…The scheme achieves state-of-the-art performance in recovery accuracy, concealment security, and invisibility. Guan et al [22] proposed the DeepMIH framework for multi-image hiding, which can realize the steganography of multiple images by utilizing the Invertible Hidden Neural Network (IHNN).…”

Section: Information Hiding Model Based On Invertible Networkmentioning

confidence: 99%

Hiding Image within Image Based on Deep Learning

Yang¹,

Zhang²,

Wu³

et al. 2022

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

Information hiding technology is a technology that transmits secret information through a carrier, and is a research hotspot in the field of information security. The traditional information hiding algorithm relies on the meticulous design of human beings, and obtains the dense image through modification. With the development of deep learning, the integration of information hiding technology and deep learning has resulted in many information hiding technologies based on deep learning. Among them, image hiding has become a research hotspot due to its large steganographic capacity. Therefore, this paper reviews the information hiding technology based on deep learning. According to the difference of hidden models, it is analysed from four aspects: (1) information hiding model based on encoder-decoder; (2) information hiding model based on generative adversarial network; (3) information hiding model based on invertible network; (4) information hiding model based on neural network information hiding models for style transfer. Finally, these models are analysed and compared, and the future development direction is discussed and prospected.

show abstract

“…Lu [ 53 ] proposed hiding multiple color images in a color image of the same size, which greatly increased the hidden capacity of image steganography. Guan [ 54 ] added an importance map module on the basis of invertible neural networks, which guided the hiding of the next image with the hidden results of the current image, avoiding the problems of contour shading and color distortion caused by multi-image hiding. This also improved the invisibility of secret information.…”

Section: Related Workmentioning

confidence: 99%

Lossless Image Steganography Based on Invertible Neural Networks

Liu

Tang

Zheng

2022

Entropy

View full text Add to dashboard Cite

Image steganography is a scheme that hides secret information in a cover image without being perceived. Most of the existing steganography methods are more concerned about the visual similarity between the stego image and the cover image, and they ignore the recovery accuracy of secret information. In this paper, the steganography method based on invertible neural networks is proposed, which can generate stego images with high invisibility and security and can achieve lossless recovery for secret information. In addition, this paper introduces a mapping module that can compress information actually embedded to improve the quality of the stego image and its antidetection ability. In order to restore message and prevent loss, the secret information is converted into a binary sequence and then embedded in the cover image through the forward operation of the invertible neural networks. This information will then be recovered from the stego image through the inverse operation of the invertible neural networks. Experimental results show that the proposed method in this paper has achieved competitive results in the visual quality and safety of stego images and achieved 100% accuracy in information extraction.

show abstract

DeepMIH: Deep Invertible Network for Multiple Image Hiding

Cited by 111 publications

References 46 publications

Smaller Is Bigger: Rethinking the Embedding Rate of Deep Hiding

Smaller Is Bigger: Rethinking the Embedding Rate of Deep Hiding

Hiding Image within Image Based on Deep Learning

Lossless Image Steganography Based on Invertible Neural Networks

Contact Info

Product

Resources

About