An image steganography scheme based on ResNet

Liu, Lianshan; Meng, Lingzhuang; Wang, Xiaoli; Peng, Yanjun

doi:10.1007/s11042-022-13206-2

Cited by 13 publications

(4 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…ResNet is one of the most widely used deep learning models in pattern and image recognition. ResNet can train a large number of data sets and store the acquired knowledge. , ResNet can obtain enough features by increasing the network depth so that it can fit nonlinear function as far as possible to achieve the purpose of approximation. , So, ResNet architecture was chosen to explore its potential for microorganism recognition.…”

Section: Resultsmentioning

confidence: 99%

Identification of Microorganism in Infected Wounds by Positively Charged Selective Sensor Array and Deep Learning Algorithm

Zhang,

Ma,

Wang

et al. 2024

Anal. Chem.

View full text Add to dashboard Cite

Microorganism are ubiquitous and intimately connected with human health and disease management. The accurate and fast identification of pathogenic microorganisms is especially important for diagnosing infections. Herein, three tetraphenylethylene derivatives (S-TDs: TBN, TPN, and TPI) featuring different cationic groups, charge numbers, emission wavelengths, and hydrophobicities were successfully synthesized. Benefiting from distinct cell wall binding properties, S-TDs were collectively utilized to create a sensor array capable of imaging various microorganisms through their characteristic fluorescent signatures. Furthermore, the interaction mechanism between S-TDs and different microorganisms was explored by calculating the binding energy between S-TDs and cell membrane/wall constituents, including phospholipid bilayer and peptidoglycan. Using a combination of the fluorescence sensor array and a deep learning model of residual network (ResNet), readily differentiation of Gram-negative bacteria (G−), Gram-positive bacteria (G+), fungi, and their mixtures was achieved. Specifically, by extensive training of two ResNet models with large quantities of images data from 14 kinds of microorganism stained with S-TDs, identification of microorganism was achieved at high-level accuracy: over 92.8% for both Gram species and antibiotic-resistant species, with 90.35% accuracy for the detection of mixed microorganism in infected wound. This novel method provides a rapid and accurate method for microbial classification, potentially aiding in the diagnosis and treatment of infectious diseases.

show abstract

Section: Resultsmentioning

confidence: 99%

Identification of Microorganism in Infected Wounds by Positively Charged Selective Sensor Array and Deep Learning Algorithm

Zhang,

Ma,

Wang

et al. 2024

Anal. Chem.

View full text Add to dashboard Cite

show abstract

“…[20] proposed a steganography method based on generative adversarial networks (GANs) for digital images, Liu at al. [21] proposed an image steganography scheme based on generalized Gaussian distribution and orthogonal matching pursuit (OMP), and Gupta and Rawat [22] proposed a novel image steganography approach based on multi-layer perceptron (MLP) and convolutional neural network (CNN). However, the work volume does not allow for a detailed discussion of these items.…”

Section: Frequency Domain Techniquesmentioning

confidence: 99%

Hiding Information in Digital Images Using Ant Algorithms

Boryczka,

Kazana

2023

Entropy

View full text Add to dashboard Cite

Stenographic methods are closely related to the security and confidentiality of communications, which have always been essential domains of human life. Steganography itself is a science dedicated to the process of hiding information in public communication channels. Its main idea is to use digital files or even communication protocols as a medium inside which data are hidden. The present research aims to investigate the applicability of ant algorithms in steganography and evaluate the effectiveness of this approach. Ant systems could be employed both in spatial and frequency-based image steganography. The combination of frequency domain and optimization method to increase robustness is used, and an integer wavelet transform is performed on the host image. ACO optimization is used to find the optimal coefficients describing where to hide the data. The other method utilizes ACO to determine the optimal pixel locations for embedding secret data in the cover image. ACO is also used to detect complex regions of the cover image. Afterward, the least-significant-bits (LSB) substitution is used to hide secret information in the detected complex regions’ pixels. Our study focuses on optimizing two mutually exclusive features of steganograms—high capacity and low distortion. An attempt was made to use ant systems to select areas of digital images that allow the greatest amount of information to be hidden with the least loss of image quality. The effect of variants of the ant system and its parameters on the quality of the results obtained was also investigated, and the final effectiveness of the proposed method was evaluated. The results of the experiments were compared with those published in related articles. The proposed procedures proved to be effective and allowed the embedding of large amounts of data with relatively little impact on image quality.

show abstract

“…Edges were used in the embedding of the image via most-significant bit (MSB) and two axes of LSB and using edges to increase the payload capacity of the image for the data, where the number of bits used in one pixel reached three bits, where the result of the embedding was efficient in terms of capacity, but few in terms of security [19]. A new method was proposed by [20] of using training for the method of hiding deep information, which proved its worth in including a large amount of data in the cover of the image, and through the outputs represented by peak signal to noise ratio (PSNR) and structural similarity index measure (SSIM), the method proved the efficiency of the method, which reflects the strength and applicability of the method used.…”

Section: Related Workmentioning

confidence: 99%

Improve steganography system using agents software based on statistical and classification technique

Halboos¹,

Al-Bakry²

2023

Bulletin EEI

View full text Add to dashboard Cite

In digital communications, information security is a paramount necessity. In the hiding algorithm, there are three basic parameters: security, capacity, and imperceptibility. Therefore, there are many ways to design the steganography algorithm, such as least significant bit (LSB), discrete wave transformation (DWT), and discrete cosine transform (DCT). The aim of this paper is to improve agent software design based on a steganography system. It proposed an agent system based on a support vector machine (SVM) classifier to hide a secret message in a certain cover image. The common dataset for steganography uses 80% training and 20% testing to get accurate results. Developing an agent system depends on six statistical parameters such as energy, standard deviation, histogram, variance, mean, and entropy. This resulted in features classified by the SVM classifier to predict the best cover image to be nominated for embedding. Worthy results were obtained in terms of imperceptibility, attack, and cover image prediction by statistical issues.

show abstract

An image steganography scheme based on ResNet

Cited by 13 publications

References 28 publications

Identification of Microorganism in Infected Wounds by Positively Charged Selective Sensor Array and Deep Learning Algorithm

Identification of Microorganism in Infected Wounds by Positively Charged Selective Sensor Array and Deep Learning Algorithm

Hiding Information in Digital Images Using Ant Algorithms

Improve steganography system using agents software based on statistical and classification technique

Contact Info

Product

Resources

About