Seek-and-Hide: Adversarial Steganography via Deep Reinforcement Learning

Pan, Wenwen; Yin, Yanling; Wang, Xinchao; Jing, Yongcheng; Song, Mingli

doi:10.1109/tpami.2021.3114555

Cited by 14 publications

(10 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Based on this analysis it can be observed that the proposed model is 6% more effective than DRL [4], 4.5% more effective than BMSE [9], and 9% more effective than GAN [11] under different test data set, which can also be observed from figure 5, wherein accuracy values are visualized. The results of PSNR are shown in figure 6, wherein PSNR of the proposed model is compared with other reference models.…”

Section: Visualmentioning

confidence: 61%

“…Where, 𝐼 𝑜𝑟𝑖𝑔 &𝐼 𝑑𝑒𝑐𝑜𝑑𝑒𝑑 represents original decoded inputs, while 𝑡 𝑐𝑜𝑚𝑝𝑙𝑒𝑡𝑒 &𝑡 𝑠𝑡𝑎𝑟𝑡 represents completion and start timestamps for the steganographic operations with 𝑁 number of values per input sets. As per these evaluations, the MMSE was compared with DRL [4], BMSE [9], and GAN [11] w.r.t. Test Data Samples (TDS) in table 1 as shown.…”

Section: ░ 4 Results and Analysismentioning

confidence: 99%

“…Spatial domain bitmap image methods transform pixels visibly, offering high capacity but potentially compromising security. RGB techniques like LSB replacement, pixel indication, optimal pixel adjustment, and secure key-based image realization minimize impact on original pixel values while inserting the message sequentially into the right-most bits of the pixel array [3,4]. Indicator channels enhance security in pixel indicator and LSB techniques [1].…”

Section: Figure 1 Design Of a Typical Image Steganographic Processmentioning

confidence: 99%

“…The effectiveness of a steganographic system hinges on its ability to fulfill all these characteristics. Researchers have been proposed numerous distinct approaches including discrete wavelet transforms (DWT), discrete cosine transforms (DCT), and least significant bit (LSB) [1,2], may be used to disguise information sets via Deep Reinforcement Learning (DRL) [3,4]. Integrating encryption with high-speed communication protocols such as multiple-input multipleoutput (MIMO) systems, orthogonal frequency division multiplexing (OFDM), and others enables secure data delivery.…”

Section: ░ 1 Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Design of an Efficient & Secure Steganographic Model using Modified LSB & Encryption Process

Yadav,

Singh

2024

IJEER

View full text Add to dashboard Cite

This paper introduces a novel steganographic model for robust multimodal data security, seamlessly integrating a modified Least Significant Bit (LSB) technique with encryption, making it applicable to diverse data types such as images, audio, video, and text. Overcoming challenges posed by existing complex models and communication delays, our approach employs a modified LSB technique to encode similar sized data samples, followed by dynamic bioinspired elliptic curve cryptography (BECC) utilizing a Mayfly Optimization (MO) Model. This adaptive strategy optimizes curve types and prime key sets, significantly enhancing data security while minimizing delays and complexities across diverse data sizes. The proposed model achieves an 8.3% reduction in encryption and steganographic process delays, while simultaneously maintaining superior Peak Signal to Noise Ratio (PSNR) and lower Mean Squared Error (MSE) levels compared to existing methods when applied to the same data samples. This highlights its effectiveness in securing dynamic datasets without compromising efficiency.

show abstract

Section: Visualmentioning

confidence: 61%

Section: ░ 4 Results and Analysismentioning

confidence: 99%

Section: Figure 1 Design Of a Typical Image Steganographic Processmentioning

confidence: 99%

Section: ░ 1 Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Design of an Efficient & Secure Steganographic Model using Modified LSB & Encryption Process

Yadav,

Singh

2024

IJEER

View full text Add to dashboard Cite

show abstract

“…To maximise the benefits from a simulated steganalytic environment, an agent in SPAR-RL used a policy network to break down the embedding process into pixel-wise actions, while the environment used an environment network to award rewards to individual pixels. An adaptive local image steganography (AdaSteg) method was presented by Pan et al (Pan et al, 2022) to enable image steganography that is both scale-and location-adaptive. The suggested technique increased the security of steganography by adaptively hiding the secret on a local scale, and it also made it possible to conceal several secrets under a single cover.…”

Section: Reinforcement Learning Based Steganographic Algorithmsmentioning

confidence: 99%

Latest Trends in Deep Learning Techniques for Image Steganography

Kumar

Sharma

Kumar

et al. 2023

International Journal of Digital Crime and Forensics

View full text Add to dashboard Cite

The development of deep convolutional neural networks has been largely responsible for the significant strides forward made in steganography over the past decade. In the field of image steganography, generative adversarial networks (GAN) are becoming increasingly popular. This study describes current development in image steganographic systems based on deep learning. The authors' goal is to lay out the various works that have been done in image steganography using deep learning techniques and provide some notes on the various methods. This study proposed a result that could open up some new avenues for future research in deep learning based on image steganographic methods. These new avenues could be explored in the future. Moreover, the pros and cons of current methods are laid out with several promising directions to define problems that researchers can work on in future research avenues.

show abstract

Enhancing the anti-steganalysis ability of steganography via adversarial examples

et al. 2023

Multimed Tools Appl

View full text Add to dashboard Cite

Seek-and-Hide: Adversarial Steganography via Deep Reinforcement Learning

Cited by 14 publications

References 41 publications

Design of an Efficient & Secure Steganographic Model using Modified LSB & Encryption Process

Design of an Efficient & Secure Steganographic Model using Modified LSB & Encryption Process

Latest Trends in Deep Learning Techniques for Image Steganography

Enhancing the anti-steganalysis ability of steganography via adversarial examples

Contact Info

Product

Resources

About