Search

Your search keyword '"Zhou, Yicong"' showing total 844 results
Start Over Author "Zhou, Yicong"
844 results on '"Zhou, Yicong"'

155. ExS-GAN: Synthesizing Anti-Forensics Images via Extra Supervised GAN

Author

Ding, Feng, Shen, Zhangyi, Zhu, Guopu, Kwong, Sam, Zhou, Yicong, and Lyu, Siwei

Abstract

So far, researchers have proposed many forensics tools to protect the authenticity and integrity of digital information. However, with the explosive development of machine learning, existing forensics tools may compromise against new attacks anytime. Hence, it is always necessary to investigate anti-forensics to expose the vulnerabilities of forensics tools. It is beneficial for forensics researchers to develop new tools as countermeasures. To date, one of the potential threats is the generative adversarial networks (GANs), which could be employed for fabricating or forging falsified data to attack forensics detectors. In this article, we investigate the anti-forensics performance of GANs by proposing a novel model, the ExS-GAN, which features an extra supervision system. After training, the proposed model could launch anti-forensics attacks on various manipulated images. Evaluated by experiments, the proposed method could achieve high anti-forensics performance while preserving satisfying image quality. We also justify the proposed extra supervision via an ablation study.

Published
2023
Full Text
View/download PDF

156. Generation of n-Dimensional Hyperchaotic Maps Using Gershgorin-Type Theorem and its Application

Author

Zhang, Yinxing, Hua, Zhongyun, Bao, Han, Huang, Hejiao, and Zhou, Yicong

Abstract

High-dimensional (HD) chaotic map has wide applications in various research fields such as neural networks and secure communication. Designing HD chaotic maps with expected dynamics and robust hyperchaotic behaviors is an interesting but challenging topic. In this article, we propose an $n$ -dimensional hyperchaotic map $(n\text{D}$ -HCM) generation method on the basis of the Gershgorin-type theorem. First, the general form of the proposed $n\text{D}$ -HCM is built using $n$ parametric polynomials. Then, the entity and coefficient parameter matrices are configured according to the Gershorin-type theorem. Theoretical analysis shows that the generated $n\text{D}$ -HCM has $n$ positive Lyapunov exponents and thus can show robust hyperchaotic behaviors. Two examples of hyperchaotic map with specified equations are provided and their properties are analyzed to show the availability of the proposed method. Performance evaluations display that our $n\text{D}$ -HCM possesses abundant properties and complex behaviors, and it can outperform some representative HD chaotic maps. Moreover, to show the application of our $n\text{D}$ -HCM, we apply it to a secure communication scheme and the experimental results exhibit that it shows much better performance than these representative HD chaotic maps in resisting transmission noise.

Published
2023
Full Text
View/download PDF

167. Multimedia Security System for Security and Medical Applications

Author

Zhou, Yicong

Abstract

This dissertation introduces a new multimedia security system for the performance of object recognition and multimedia encryption in security and medical applications. The system embeds an enhancement and multimedia encryption process into the traditional recognition system in order to improve the efficiency and accuracy of object detection and recognition while protecting multimedia data. The dissertation then focuses on the system's enhancement and encryption processes, and presents contributions to the area of image enhancement and multimedia encryption. For the purposes of image enhancement, a new enhancement measure called the second-derivative-like measure of enhancement (SDME) is introduced to quantitatively evaluate the enhancement algorithm's performance. This is followed by the introduction of a new 3D CT baggage image enhancement algorithm for homeland security applications. Computer simulations and comparisons demonstrate that the presented algorithm significantly improves the visual quality of objects in the original CT images while reducing background noise. The quantitative SDME measure results and 3D visualizations verify the algorithm's excellent enhancement performance. To improve the visual quality of medical images and thereby improve the efficiency and accuracy of early cancer detection, a new nonlinear filter, called the Alpha-Weighted Quadratic Filter (AWQF), is integrated with human visual system based decomposition and unsharp masking techniques to enhance mammograms for breast cancer detection. To enhance prostate MR images for prostate cancer detection, the same filter is integrated with alpha-trimmed mean separation and logarithmic enhancement techniques. Simulation results and comparisons are given to demonstrate the enhancement algorithms' excellent performance when it comes to enhancing mammograms and prostate MR images. In order to enhance the efficiency and security of existing encryption algorithms, five parametric recursive sequences and their transforms are developed and applied to multimedia scrambling/encryption. The recursive sequence transforms are integrated with the Fibonacci bit-plane decomposition method and the newly introduced (" n, k, p")-Gray code bit-plane decomposition for image and object encryption. Based on the concept of using one set of security keys to encrypt the original image and a different set of security keys to reconstruct the image to obtain the final encrypted image, a discrete parametric cosine transform is used for image encryption. Despite the fact that the edge map has been used traditionally for image compression, enhancement, and recognition, it has never been used for image encryption. This dissertation investigates the use of the edge map for applications in image encryption. The edge map is integrated with the 3D Cat Map for image encryption and with the chaotic logistic map for encrypting medical images for the sake of privacy protection. The concept of the edge map is then extended further to produce a binary "key-image", which can either be a bit-plane or an edge map obtained from any other image. Computer simulations, comparisons and security analysis demonstrate the excellent performance of the presented encryption algorithms when it comes to multimedia encryption. The contributions of this dissertation are as follows: (1) A new multimedia security system for object detection and multimedia encryption (2) Image enhancement: (a) A new SDME measure for quantitatively evaluating the enhancement algorithm's performance (b) A new 3D CT baggage image enhancement algorithm using alpha weighted mean separation for homeland security applications (c) A new alpha weighted quadratic filter for mammogram enhancement (d) A new human visual system decomposition based algorithm for mammogram enhancement (e) A new nonlinear unsharp masking scheme for mammogram enhancement (f) A new prostate MR image enhancement algorithm using the alpha-trimmed mean separation and nonlinear filtering (g) A new logarithmic enhancement algorithm using nonlinear filtering for prostate MR image enhancement (3) Multimedia encryption: (a) Five new parametric recursive sequences and their transforms for multimedia encryption, including the truncated P-Fibonacci sequence, P-Lucas sequence, ("n", "k," "p")--Gray code, P-recursive sequence and the parametric M-sequence. (b) A new 2D P-recursive transform (c) Two P-recursive transform based multimedia encryption algorithms in the spatial and frequency domains (d) A new truncated Fibonacci p-code bit-plane decomposition (e) A new ("n," "k," "p")--Gray code bit-plane decomposition (f) A new image encryption algorithm using the P-Fibonacci transform and decomposition (g) A new selective object encryption algorithm using the truncated Fibonacci p-code bit-plane decomposition (h) A new image encryption algorithm using the ("n," "k," "p")--Gray code and its decomposition (i) A new image encryption algorithm using the Discrete Parametric Cosine Transform (j) A new image encryption algorithm using the edge map and a new 3D Cat Map (k) A new medical image encryption algorithm using the edge map and chaotic logistic map (l) Two image encryption algorithms using binary key-images: one uses bit-plane and the other uses the edge map. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

Published
2010

181. Two‐order graph convolutional networks for semi‐supervised classification

Author

Fu Sichao, Zhou Yicong, Li Shuying, and Liu Weifeng

Subjects
Approximation theory, Artificial neural network, business.industry, Computer science, Deep learning, 020206 networking & telecommunications, Graph theory, Pattern recognition, 02 engineering and technology, Convolutional neural network, Signal Processing, 0202 electrical engineering, electronic engineering, information engineering, Graph (abstract data type), 020201 artificial intelligence & image processing, Computer Vision and Pattern Recognition, Artificial intelligence, Electrical and Electronic Engineering, business, Laplace operator, Software
Abstract

Currently, deep learning (DL) algorithms have achieved great success in many applications including computer vision and natural language processing. Many different kinds of DL models have been reported, such as DeepWalk, LINE, diffusionconvolutional neural networks, graph convolutional networks (GCN), and so on. The GCN algorithm is a variant of convolutional neural network and achieves significant superiority by using a one-order localised spectral graph filter. However, only a one-order polynomial in the Laplacian of GCN has been approximated and implemented, which ignores undirect neighbour structure information. The lack of rich structure information reduces the performance of the neural networks in the graph structure data. In this study, the authors deduce and simplify the formula of two-order spectral graph convolutions to preserve rich local information. Furthermore, they build a layerwise GCN based on this two-order approximation, i.e. two-order GCN (TGCN) for semi-supervised classification. With the two-order polynomial in the Laplacian, the proposed TGCN model can assimilate abundant localised structure information of graph data and then boosts the classification significantly. To evaluate the proposed solution, extensive experiments are conducted on several popular datasets including the Citeseer, Cora, and PubMed dataset. Experimental results demonstrate that the proposed TGCN outperforms the state-of-art methods.

Published
2019

Catalog

Books, media, physical & digital resources
See catalog results