Applications of Machine Learning in Steganography for Data Protection and Privacy

2021 ◽  
pp. 306-325
Author(s):  
Mahip M. Bartere ◽  
Sneha Bohra ◽  
Prashant Adakane ◽  
B. Santhosh Kumar
Keyword(s):  
Machine Learning ◽  
Data Hiding ◽  
Data Protection ◽  
Data Security ◽  
Data Extraction ◽  
Original Data ◽  
Extraction Techniques ◽  
Reverse Engineer ◽  
Applications Of Machine Learning

Data security is one of the most important aspects in today's scenario. Whenever we send our data from source to destination, data protection is one of the prime components. With the help of data hiding and data extraction techniques, we are able to provide the solution of different types of problems whenever we transfer our data. Steganography is a process where we can hide our data and maintain the quality of the image. At the same time, we think about data alteration. With the help of stegtanalysis method, we reverse engineer and extract the original data. In this chapter, data hiding and data extraction techniques are explained in the combination of machine learning architecture. The combination of steganography and steganalysis along with machine learning is used to identify protected data using different techniques.

scholarly journals Developing audio data hiding scheme using random sample bits with logical operators

2019 ◽  
Vol 13 (1) ◽  
pp. 147 ◽  
Author(s):  
Mohammed Hatem Ali Al-Hooti ◽  
Tohari Ahmad ◽  
Supeno Djanali
Keyword(s):  
Information Technology ◽  
Random Sample ◽  
Data Hiding ◽  
Data Security ◽  
Secret Data ◽  
Audio File ◽  
Logical Operators ◽  
Audio Data ◽  
Case Data

Sharp development progress of information technology has affected many aspects including data security. This is because classified data are often transferred between systems. In this case, data hiding exists to protect such data. Some methods which have been proposed, however, are not yet optimal concerning the amount of the secret and the quality of the resulted stego data. In this paper, we explore an audio file as the medium to carry the secret data which has been extracted into binary. Before the process begins, the cover is converted to binary and each sample’s bits are divided into two groups, one is used as the location of the embedded 4 bits whereas the second part locates the two bits that are randomly selected as the key. The experimental results have validated that the capacity is high and there is no much impact on the quality. Moreover, compared to the current LSB methods the security is exceedingly enhanced.

scholarly journals Classification of masked image data

PLoS ONE ◽  
2021 ◽  
Vol 16 (7) ◽  
pp. e0254181
Author(s):  
Kamila Lis ◽  
Mateusz Koryciński ◽  
Konrad A. Ciecierski
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Image Data ◽  
Original Data ◽  
Machine Learning Algorithms ◽  
General Data Protection Regulation ◽  
Additional Information ◽  
Classification Of Images ◽  
Applications Of Machine Learning

Data classification is one of the most commonly used applications of machine learning. The are many developed algorithms that can work in various environments and for different data distributions that perform this task with excellence. Classification algorithms, just like other machine learning algorithms have one thing in common: in order to operate on data, they must see the data. In the present world, where concerns about privacy, GDPR (General Data Protection Regulation), business confidentiality and security are growing bigger and bigger; this requirement to work directly on the original data might become, in some situations, a burden. In this paper, an approach to the classification of images that cannot be directly accessed during training has been made. It has been shown that one can train a deep neural network to create such a representation of the original data that i) without additional information, the original data cannot be restored, and ii) that this representation—called a masked form—can still be used for classification purposes. Moreover, it has been shown that classification of the masked data can be done using both classical and neural network-based classifiers.

scholarly journals High Complexity Bit-Plane Security Enchancement in BPCS Steganography

2017 ◽  
Author(s):  
Andysah Putera Utama Siahaan
Keyword(s):  
Data Hiding ◽  
Data Security ◽  
Original Data ◽  
Cover Image ◽  
Security Level ◽  
High Complexity ◽  
High Security Level ◽  
Bit Plane ◽  
Exclusive Or ◽  
Plane Form

In BPCS Steganography, data hiding will be split into blocks that have a high complexity where the blocks are categorized into informative and noise-like regions. A noise-like region is a bit-plane that has the greatest probability as a data hiding since it has a high complexity. In this region, the data inserted is vulnerable to attack. Someone can easily take a series of characters that are stored on a noise-like region previously if the system is not modified. Improving the bit-plane composition is to increase data security. Bit-plane will be combined with a specified key. The key should be changed to bit-plane form as well. The key that has already been turned into the bit-plane will be mated with the original data. Using an exclusive-or of this part is the best way to produce the cipher bit-plane. Finally, the data residing on the cover image produced have a high-security level.

scholarly journals Wine Quality Prediction Using Machine Learning

2021 ◽  
Vol 9 (8) ◽  
pp. 1997-2001
Author(s):  
Vrusha P. Sangodkar
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Feature Selection ◽  
Bp Neural Network ◽  
Data Extraction ◽  
Backpropagation Neural Network ◽  
Wine Quality ◽  
Machine Learning Methods ◽  
Culture Consumption

Abstract: Nowadays people are living a luxurious lifestyle, wine has become a part of one's culture. consumption of wine is very common throughout the world so its quality is very important. hence its important to analyse wine quality quality of the wines are usually checked by humans through tasting but it has other physicochemical attributes which affects the taste but the process is slow hence machine learning methods can be used for the same. dataset is taken and feature selection is done using pca feature selection and then accuracy is find using SVM, backpropagation neural network and Random forest algorithm to find which model fits best and gives greater accuracy. Keywords: Data Extraction, PCA, SVM,BP neural network, Randomforest

scholarly journals Reversible Data Hiding in Encrypted Image Based on (7, 4) Hamming Code and UnitSmooth Detection

Entropy ◽  
2021 ◽  
Vol 23 (7) ◽  
pp. 790
Author(s):  
Lin Li ◽  
Chin-Chen Chang ◽  
Chia-Chen Lin
Keyword(s):  
Data Hiding ◽  
Privacy Protection ◽  
Reversible Data Hiding ◽  
Signal To Noise Ratio ◽  
Data Extraction ◽  
Hamming Code ◽  
Hiding Capacity ◽  
Encrypted Images

With the development of cloud storage and privacy protection, reversible data hiding in encrypted images (RDHEI) plays the dual role of privacy protection and secret information transmission. RDHEI has a good application prospect and practical value. The current RDHEI algorithms still have room for improvement in terms of hiding capacity, security and separability. Based on (7, 4) Hamming Code and our proposed prediction/ detection functions, this paper proposes a Hamming Code and UnitSmooth detection based RDHEI scheme, called HUD-RDHEI scheme for short. To prove our performance, two database sets—BOWS-2 and BOSSBase—have been used in the experiments, and peak signal to noise ratio (PSNR) and pure embedding rate (ER) are served as criteria to evaluate the performance on image quality and hiding capacity. Experimental results confirm that the average pure ER with our proposed scheme is up to 2.556 bpp and 2.530 bpp under BOSSBase and BOWS-2, respectively. At the same time, security and separability is guaranteed. Moreover, there are no incorrect extracted bits during data extraction phase and the visual quality of directly decrypted image is exactly the same as the cover image.

scholarly journals Abduction-Based Explanations for Machine Learning Models

2019 ◽  
Vol 33 ◽  
pp. 1511-1519 ◽  
Author(s):  
Alexey Ignatiev ◽  
Nina Narodytska ◽  
Joao Marques-Silva
Keyword(s):  
Machine Learning ◽  
Decision Problem ◽  
Abductive Reasoning ◽  
Decision Makers ◽  
Reasoning System ◽  
Constraint Reasoning ◽  
Human Decision ◽  
Applications Of Machine Learning ◽  
Machine Learning Models

The growing range of applications of Machine Learning (ML) in a multitude of settings motivates the ability of computing small explanations for predictions made. Small explanations are generally accepted as easier for human decision makers to understand. Most earlier work on computing explanations is based on heuristic approaches, providing no guarantees of quality, in terms of how close such solutions are from cardinality- or subset-minimal explanations. This paper develops a constraint-agnostic solution for computing explanations for any ML model. The proposed solution exploits abductive reasoning, and imposes the requirement that the ML model can be represented as sets of constraints using some target constraint reasoning system for which the decision problem can be answered with some oracle. The experimental results, obtained on well-known datasets, validate the scalability of the proposed approach as well as the quality of the computed solutions.

scholarly journals Trends in Development of Novel Machine Learning Methods for the Identification of Gliomas in Datasets That Include Non-Glioma Images: A Systematic Review

2021 ◽  
Vol 11 ◽  
Author(s):  
Harry Subramanian ◽  
Rahul Dey ◽  
Waverly Rose Brim ◽  
Niklas Tillmanns ◽  
Gabriel Cassinelli Petersen ◽  
...  
Keyword(s):  
Machine Learning ◽  
Systematic Review ◽  
Clinical Practice ◽  
Data Extraction ◽  
Final Analysis ◽  
Poor Quality ◽  
Controlled Vocabulary ◽  
Quality Of Reporting ◽  
Testing Strategies

PurposeMachine learning has been applied to the diagnostic imaging of gliomas to augment classification, prognostication, segmentation, and treatment planning. A systematic literature review was performed to identify how machine learning has been applied to identify gliomas in datasets which include non-glioma images thereby simulating normal clinical practice.Materials and MethodsFour databases were searched by a medical librarian and confirmed by a second librarian for all articles published prior to February 1, 2021: Ovid Embase, Ovid MEDLINE, Cochrane trials (CENTRAL), and Web of Science-Core Collection. The search strategy included both keywords and controlled vocabulary combining the terms for: artificial intelligence, machine learning, deep learning, radiomics, magnetic resonance imaging, glioma, as well as related terms. The review was conducted in stepwise fashion with abstract screening, full text screening, and data extraction. Quality of reporting was assessed using TRIPOD criteria.ResultsA total of 11,727 candidate articles were identified, of which 12 articles were included in the final analysis. Studies investigated the differentiation of normal from abnormal images in datasets which include gliomas (7 articles) and the differentiation of glioma images from non-glioma or normal images (5 articles). Single institution datasets were most common (5 articles) followed by BRATS (3 articles). The median sample size was 280 patients. Algorithm testing strategies consisted of five-fold cross validation (5 articles), and the use of exclusive sets of images within the same dataset for training and for testing (7 articles). Neural networks were the most common type of algorithm (10 articles). The accuracy of algorithms ranged from 0.75 to 1.00 (median 0.96, 10 articles). Quality of reporting assessment utilizing TRIPOD criteria yielded a mean individual TRIPOD ratio of 0.50 (standard deviation 0.14, range 0.37 to 0.85).ConclusionSystematic review investigating the identification of gliomas in datasets which include non-glioma images demonstrated multiple limitations hindering the application of these algorithms to clinical practice. These included limited datasets, a lack of generalizable algorithm training and testing strategies, and poor quality of reporting. The development of more robust and heterogeneous datasets is needed for algorithm development. Future studies would benefit from using external datasets for algorithm testing as well as placing increased attention on quality of reporting standards.Systematic Review Registrationwww.crd.york.ac.uk/prospero/display_record.php?ID=CRD42020209938, International Prospective Register of Systematic Reviews (PROSPERO 2020 CRD42020209938).

scholarly journals Radiomics for Precision Medicine in Glioblastoma

2021 ◽  
Author(s):  
Kiran Aftab ◽  
Faiqa Binte Aamir ◽  
Saad Mallick ◽  
Fatima Mubarak ◽  
Whitney B. Pope ◽  
...  
Keyword(s):  
Machine Learning ◽  
Data Extraction ◽  
Primary Brain Tumor ◽  
Treatment Modalities ◽  
Patient Specific ◽  
Extraction Techniques ◽  
Non Invasive ◽  
Mri Sequences ◽  
Larger Sample

Abstract Introduction: Being the most common primary brain tumor, glioblastoma presents as an extremely challenging malignancy to treat with dismal outcomes despite treatment. Varying molecular epidemiology of glioblastoma between patients and intra-tumoral heterogeneity explains the failure of current one-size-fits-all treatment modalities. Radiomics uses machine-learning to identify salient features of the tumor on brain imaging and promises patient specific management in glioblastoma patients. Methods: We performed a comprehensive review of the available literature on studies investigating the role of radiomics and radiogenomics models for the diagnosis, stratification, prognostication as well as treatment planning and monitoring of glioblastoma.Results: Classifiers based on combination of various MRI sequences, genetic information and clinical data can predict non-invasive tumor diagnosis, overall survival and treatment response with reasonable accuracy. However, the use of radiomics for glioblastoma treatment remains in infancy as larger sample sizes, standardized image acquisition and data extraction techniques are needed to develop machine learning models that can be translated effectively into clinical practice. Conclusion: Radiomics has the potential to transform the scope of glioblastoma management through personalized medicine.

scholarly journals Peningkatan Kualitas Citra Stego pada Adaptive Pixel Block Grouping Reduction Error Expansion dengan Variasi Model Scanning pada Pembentukan Kelompok Piksel

2018 ◽  
Vol 5 (2) ◽  
pp. 185
Author(s):  
Hendro Eko Prabowo ◽  
Tohari Ahmad
Keyword(s):  
Data Hiding ◽  
Data Security ◽  
Secret Data ◽  
Stego Image ◽  
Error Expansion ◽  
Data Adaptive ◽  
Reduction Error ◽  
Carrier Image ◽  
Column Base

<p class="Abstrak">ebutuhan komunikasi yang terus bertambah dan ditandai dengan meningkatnya jumlah <em>IP traffic</em> dari 744 EB menjadi 1.164 EB menjadikan keamanan sebagai salah satu kebutuhan utama dalam menjaga kerahasiaan data. <em>Adaptive Pixel Block Grouping Reduction Error Expansion (APBG-REE)</em> sebagai salah satu metode data hiding dapat diterapkan untuk memenuhi kebutuhan tersebut. Metode ini akan membagi citra carrier menjadi blok-blok dan membentuknya menjadi kelompok-kelompok piksel. Hasil dari proses ini akan dimanfaatkan untuk menyembunyikan data rahasia. Namun, metode ini memiliki kekurangan, yaitu belum diketahuinya metode <em>scanning</em> terbaik dalam pembentukan kelompok piksel untuk menciptakan citra <em>stego</em> dengan kualitas tinggi. Untuk mengatasi masalah ini, kami mengusulkan 4 mode (cara) <em>scanning</em> berdasarkan arah <em>scanning</em> tersebut. Mode <em>scanning</em> tersebut memberikan hasil yang berbeda-beda untuk masing-masing citra <em>stego</em> yang diujikan. Namun berdasarkan hasil uji coba, setiap mode <em>scanning</em> mampu menjaga kualitas citra stego diatas 57,5 dB. Hasil ini akan meningkat seiring dengan berkurangnya jumlah <em>shifted pixel</em> yang terbentuk.</p><p class="Abstrak"> </p><p class="Abstrak"><em><strong>Abstract</strong></em></p><p class="Abstrak"><em>The need of communication has increased continously which is represented by the rise of number of IP traffic, from 744 EB to 1.164 EB. This has made data security one of the main requirements in terms of securing secret data. Adaptive Pixel Block Grouping Reduction Error Expansion (APBG-REE) as one of data hiding methods can be implemented to meet that requirement. It divides the carrier image into blocks which are then used as pixel groups. The result of this process is to be a space for secret data. However, this method has a problem in the scanning when creating pixel groups to generate a high quality stego image. To handle this problem, we propose four scanning models base on its direction. This means that the scanning can be done row-by-row or column-by-column. Base on the experiment, we find that those modes deliver various results and each of them is able to maintain the stego quality of more than 57,5 dB. This result increases along with the decreasing the number of shifted pixels.</em></p>

Sign in / Sign up

Export Citation Format

Share Document