Online Sequential Extreme Learning Machine for watermarking in DWT domain

Singh, Ram; Dabas, Neelam; Chaudhary, Vidhi; Nagendra,

doi:10.1016/j.neucom.2015.03.115

Cited by 23 publications

(10 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Furthermore, a scaling factor is assigned for embedding strength in each block which is estimated by genetic algorithm. A similar embedding method is utilized by [16], which propose Extreme Learning Machine (ELM) for prediction in the DWT domain. In spite of vast research proposals for the application of machine learning tools in watermarking, none of the above-mentioned methods propose a unified watermarking framework based on machine learning approaches.…”

Section: Related Workmentioning

confidence: 99%

“…Nowadays the application of machine learning tools in watermarking is growing very rapidly, because of their effective solutions to embedding and extraction processes [12,13,14,15,16,17]. Nevertheless, most of them generally utilize machine learning tools such as Support Vector Machine (SVM) [18], Support Vector Regression (SVR) [19], Radial Basic Function Neural Network (RBFNN) [20], and K Nearest Neighbor (KNN) [21] for specific parts of watermarking procedure such as parameter optimization [12,13], prediction of transform domain coefficients [14,15,16] and attack estimation several image blocks, rather than simply swapping in a single block. Thus, the watermarked image demonstrates impressive robustness against several heavy attacks.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

ReDMark: Framework for residual diffusion watermarking based on deep networks

Ahmadi

Norouzi

Karimi

et al. 2020

Expert Systems with Applications

177

122

View full text Add to dashboard Cite

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

ReDMark: Framework for residual diffusion watermarking based on deep networks

Ahmadi

Norouzi

Karimi

et al. 2020

Expert Systems with Applications

177

122

View full text Add to dashboard Cite

“…Compared with the multiple classifier ensemble algorithms, the single classifier learning algorithms have lower computational complexity but poor adaptability. The typical algorithms include VFDT [15], ELM-TV [16], ODTC [17], iCaRL [18] and so on. The multiple classifier algorithms are more popular than the single classifier algorithms.…”

Section: The Related Research Workmentioning

confidence: 99%

A Parallel and Reverse Learn++.NSE Classification Algorithm

Shen

Tong

et al. 2020

IEEE Access

View full text Add to dashboard Cite

There are lots of typical applications of classification learning for accumulated big data in the nonstationary environments. It is very necessary and urgent to study the algorithms that can carry out classification learning efficiently in these environments. The recently proposed algorithm, named Learn++.NSE, has made an important breakthrough, which is one of the important research achievements in this research field. However, the Learn++.NSE algorithm adopts a serial ensemble mechanism, and its execution efficiency needs to be further improved when facing the long-term accumulated big data. A Parallel and Reverse Learn++.NSE algorithm, abbreviated as PRLearn++.NSE, is proposed in this paper by changing the ensemble mechanism of the base-classifiers, which uses the old base-classifiers as a supplement to the new base-classifier. It constructs a fast and parallel ensemble mechanism. The experimental results on the artificially generated dataset and real dataset show that the proposed PRLearn++.NSE algorithm can greatly improve the efficiency of ensemble classification learning under the premise of obtaining the approaching classification accuracy of Learn++.NSE algorithm and it is very suitable for fast ensemble classification learning for the long-term accumulated big data.

show abstract

“…Görüntü uzayı tekniklerinde damga ekleme işlemleri görüntünün gri-seviye piksel değerleri değiştirilerek gerçekleştirilir [7][8][9]. Ayrık Fourier Dönüşümü [10][11][12], Ayrık Kosinüs Dönüşümü (AKD) [13][14][15][16] ve Ayrık Dalgacık Dönüşümü (ADD) [17][18][19] gibi frekans uzayı damgalama tekniklerinde dönüşüm katsayıları değiştirilerek damga ekleme işlemleri gerçekleştirilir. Vektör kuantalama tabanlı tekniklerde, damga vektör kuantalama yöntemleriyle sıkıştırılarak vektör kuantalama uzayında orijinal görüntüye eklenebilir [20][21][22].…”

Section: Gi̇ri̇ş (Introduction)unclassified

Evaluation of Different Frequency Regions in Multiple Image Watermarking Method

Abdul-Rahman¹,

Öztürk²

2018

Bilişim Teknolojileri Dergisi

View full text Add to dashboard Cite

Özet-Son yıllarda internet ve mobil teknolojilerin gelişmesi ile birlikte, çoklu-ortam dokümanlarının telif hakkının korunması önem kazanmıştır. Sayısal görüntü ve video gibi çoklu-ortam dokümanlarının korunmasında sayısal damgalama en çok kullanılan yöntem olmuştur. Bu çalışmada, Ayrık Kosinüs Dönüşümü, Ayrık Dalgacık Dönüşümü (ADD) ve Arnold Dönüşümü kullanılarak gri-seviyeli görüntü damgalama yöntemi önerilmiştir. Bu damgalama yönteminde eklenecek damgalar dört eşit parçalara ayrılmıştır. Saldırılara karşı damgalama yönteminin dayanıklılığını artırmak için bu parçaların her biri damga eklenecek görüntülerin ADD işlemi sonucunda elde edilen dört ayrı bantlarının farklı alt bölgelerine gizlenmiştir. ADD frekans uzayında farklı alt bölgelere damga ekleme işleminde önerilen griseviyeli damgalama yönteminin başarımı damgalanmış görüntülere doğrusal ve doğrusal olmayan saldırılar uygulanarak değerlendirilmiştir. Bu çalışmada ayrıca önerilen yöntem ile temel alçak frekans bandına damga ekleme yöntemi karşılaştırılmıştır. Deneysel sonuçlar önerilen yöntemin saldırılara karşı dayanıklı olduğunu göstermiştir. Bununla birlikte damgalanmış görüntülerin kalitelerinin korunduğu da görülmüştür.Abstract-Due to the development of the internet and mobile technologies in recent years, the protection of the copyright of multimedia documents has become importance. Digital watermarking is the most widely used method for protecting multimedia documents such as digital image and video. In this study, gray-level image watermarking method is proposed using Discrete Cosine Transform, Discrete Wavelet Transform (DWT) and Arnold Transform. In this watermarking method, the watermarks to be embedded are segmented into four equal-sized parts. In order to increase the robustness of watermarking method against the attacks, each one of these parts have been embedded inside different sub-regions of the four separate bands that have been obtained as a result of DWT processing of the cover images. The performance of the proposed gray-level watermarking method for embedding watermarks in different sub regions in DWT frequency domain has been evaluated by applying linear and nonlinear attacks to the watermarked images. Also, in this study the proposed method is compared with the basic low frequency band watermarking method. Experimental results have shown that the proposed method is robust to the attacks. It is also observed that the quality of the watermarked images has been preserved.

show abstract

Online Sequential Extreme Learning Machine for watermarking in DWT domain

Cited by 23 publications

References 33 publications

ReDMark: Framework for residual diffusion watermarking based on deep networks

ReDMark: Framework for residual diffusion watermarking based on deep networks

A Parallel and Reverse Learn++.NSE Classification Algorithm

Evaluation of Different Frequency Regions in Multiple Image Watermarking Method

Contact Info

Product

Resources

About