Jinwei Wang

• Professor, PhD
• Head of Department at Nanjing University of Information Science and Technology

Detection of color images that have undergone double compression is a critical aspect of digital image forensics. Despite the existence of various methods capable of detecting double Joint Photographic Experts Group (JPEG) compression, they are unable to address the issue of mixed double compression resulting from the use of different compression s...

In recent years, various adversarial defense methods have been proposed to improve the robustness of deep neural networks. Adversarial training is one of the most potent methods to defend against adversarial attacks. However, the difference in the feature space between natural and adversarial examples hinders the accuracy and robustness of the mode...

Invisible watermarking can be used as an important tool for copyright certification in the Metaverse. However, with the advent of deep learning, Deep Neural Networks (DNNs) have posed new threats to this technique. For example, artificially trained DNNs can perform unauthorized content analysis and achieve illegal access to protected images. Furthe...

Deep neural networks (DNNs) can be susceptible to subtle perturbations that may mislead the model. While adversarial attacks are successful in the white-box setting, they are less effective in the black-box setting. To address this issue, we propose an attack method that simulates a smoothed loss function by sampling from a Gaussian distribution. W...

Detection of double Joint photographic experts group (JPEG) compression is a major part of image forensics, especially double JPEG compression detection with the same quantization matrix is still a challenging problem. Detection of double JPEG compression for different quantization matrices has achieved good performance so far, and some methods for...

With the rapid development and application of Internet technology in recent years, the issue of information security has received more and more attention. Digital steganography is used as a means of secure communication to hide information by modifying the carrier. However, steganography can also be used for illegal acts, so it is of great signific...

Deep learning models have enabled significant performance improvements to remote sensing image processing. Usually, a large number of training samples is required for detection models. In this study, a dynamic simulation training strategy is designed to generate samples in real time during training. The few adversarial examples are not only directl...

Deep Neural Networks (DNNs) are vulnerable to adversarial attacks, which makes adversarial attacks serve as a method to evaluate the robustness of DNNs. However, adversarial attacks have the disadvantage of high white-box attack success rates but low transferability. Therefore, many methods were proposed to improve the transferability of adversaria...

With the rapid proliferation of urbanization, massive data in social networks are collected and aggregated in real time, making it possible for criminals to use images as a cover to spread secret information on the Internet. How to determine whether these images contain secret information is a huge challenge for multimedia computing security. The s...

Rule-based fine-grained IP geolocation methods are hard to generalize in computer networks which do not follow hypothetical rules. Recently, deep learning methods, like multi-layer perceptron (MLP), are tried to increase generalization capabilities. However, MLP is not so suitable for graph-structured data like networks. MLP treats IP addresses as...

The quantization step is a crucial parameter in JPEG compression, that can reveal the compression history of a JPEG image. Estimating the quantization steps for single compressed and recompressed images is attracting considerable interest in the field of image forensics and steganalysis. Several effective methods have been proposed, but the perform...

Detection of aligned double Joint Photographic Experts Group (JPEG) compressed images is a crucial area of research within the field of digital image forensics. The detection tasks for aligned double JPEG compression can be categorized into two sub-tasks, namely detecting double JPEG images with the same quantization matrix (DJSQM) or double JPEG i...

Wind data play a crucial role in the advanced planning, design, and maintenance of wind farms, scheduling, energy production, and accurate distribution. Most of the measured wind dataset comprises missing observations due to several reasons, including the failure of measuring devices, human mistakes, environmental impact, and natural calamities. Co...

The rapid development of the generative adversarial networks (GANs) has made it an unprecedented success in image generation. The emergence of BigGAN, StyleGAN, and other advanced GAN makes the generated images more real and deceptive, which poses potential threats to national security, social stability, and personal privacy. In this paper, we prop...

The popularity of image editing techniques and user-friendly editing software have seriously reduced the authenticity of the images. Detection and localization of image manipulations are becoming urgent problems to be solved. Although many existing solutions attempt to address these problems, most works can only solve one specific type of manipulat...

Recently, convolutional neural networks (CNN) have shown significant success in image classification tasks. However, studies show that neural networks are susceptible to tiny perturbations. When the disturbing image is input, the neural network will make a different judgment. At present, most studies use the negative side of perturbation to mislead...

The existing detection methods of algorithmically generated malicious domain names lack theoretical modelling methods for domain name element composition. To address this problem, a semantic element representation model for domain names is constructed based on the set of semantic elements of domain names and the probabilistic context free grammar m...

Malicious intelligent algorithms greatly threaten the security of social users' privacy by detecting and analyzing the uploaded photos to social network platforms. The destruction to DNNs brought by the adversarial attack sparks the potential that adversarial examples serve as a new protection mechanism for privacy security in social networks. Howe...

Breast cancer is one of the most common types of cancer around the world. The early‐stage recognition of breast cancer is favourable for the diagnosis and treatment of the affecting patient. Data mining approaches can ease the diagnosis of breast cancer by analysis of the associated dataset for real‐time decision‐making. The present study proposes...

As Joint Photographic Experts Group (JPEG) compression is widely used in image processing and computer vision, the detection of JPEG forgery has become an important issue in digital image forensics, and many related works have been reported. However, these works all assume that the quantization table in the JPEG header file is real, and correspondi...

Deep neural networks are vulnerable to adversarial examples, which can fool classifiers by adding small perturbations. Various adversarial attack methods have been proposed in the past several years, and most of them add the perturbation in a “sparse” or “global” way. Since the number of pixels perturbed by the “sparse" method and the perturbation...

Nowadays, users upload numerous photos to social network platforms to share their daily lives. These photos contain numerous personal information, which can be easily captured by intelligent algorithms. To improve privacy security, we aim to form a protection mechanism by exploiting adversarial examples, which can mislead and disrupt intelligent al...

Although the current generative adversarial networks (GAN)-generated face forensic detectors based on deep neural networks (DNNs) have achieved considerable performance, they are vulnerable to adversarial attacks. In this paper, an effective local perturbation generation method is proposed to expose the vulnerability of state-of-the-art forensic de...

Most studies use the negative effect of perturbation to mislead the neural network, such as adversarial examples, while ignoring the positive effect of improving neural networks’ performance. In this work, we use enhanced samples with positive perturbation to improve target classifiers’ performance and propose an algorithm of strong generative adve...

Due to the popularity of JPEG format images in recent years, JPEG images will inevitably involve image editing operation. Thus, some tramped images will leave tracks of Non-aligned double JPEG ( NA-DJPEG ) compression. By detecting the presence of NA-DJPEG compression, one can verify whether a given JPEG image has been tampered with. However, only...

Recently, Zhou et al. designed a two-stream faster Region-Convolutional neural networks (R-CNN) model RGB-N for color image splicing localization in CVPR2018. However, the RGB-N locates spliced regions only at block-level and ignores the entirety and inherent correlation of three channels. Therefore, an improved quaternion two-stream R-CNN model is...

Machine learning approaches have a valuable contribution in improving competency in automated decision systems. Several machine learning approaches have been developed in the past studies in individual disease diagnosis prediction. The present study aims to develop a hybrid machine learning approach for diagnosis predictions of multiple diseases ba...

Joint photographic experts group (JPEG) compression is widely used in image processing and computer vision. Detecting double compressed JPEG images is a common problem in forensics and detecting compressed images with the same quantization matrix remains a challenging task. However, most existing methods were designed for detection in grayscale ima...

In order to overcome the disadvantages of convolution neural network (CNN) in the current hyperspectral image (HSI) classification/segmentation methods, such as the inability to recognize the rotation of spatial objects, the difficulty to capture the fine spatial features and the problem that principal component analysis (PCA) ignores some importan...

Detecting malicious domain names generated by domain generation algorithms is critical for defending the network against sophisticated attacks. In the past decade, deep-learning-based detection schemes have proven to be the most effective. However, each of these schemes requires sufficient computation time, which makes it difficult for real-time on...

Cloud services have become an increasingly popular solution to provide different services to clients. More and more data are outsourced to the cloud for storage and computing. With this comes concern about the security of outsourced data. In recent years, homomorphic encryption, blockchain, steganography, and other technologies have been applied to...

Nowadays with the development of photosensitive seal technology, the seal fraud events have gradually increased. Forged seals can bring considerable benefits to counterfeiters, and will also bring huge losses to companies and users. Since it is almost impossible to collect enough forged seal samples, traditional machine learning methods do not work...

Since the emergence of adversarial examples brings great security threat to deep neural network which is widely used in various fields, their forensics become very important. In this paper, a lightweight model for the forensics of adversarial example based on DCT-like domain is proposed. The DCT-like layer realizes the block conversion of data from...

With the development of information security, localization of image manipulations havs become a hot topic. In this paper, a hybrid loss network is proposed for the manipulated image forensics. First, the patch prediction module extracts corresponding features representing the discrepancy of the tampered region boundaries. Then, these features are c...

Currently, the demand for covert communication in open social networks brings new opportunities and challenges to existing image steganography technology in terms of robustness and security. To this end, an image robust adaptive steganography is proposed with robustness against multiple image processing attacks and detection resistance. First, a ro...

The emergence of adversarial examples has had a great impact on the development and application of deep learning. In this paper, a novel convolutional neural network model, the stochastic multifilter statistical network (SmsNet), is proposed for the detection of adversarial examples. A feature statistical layer is constructed to collect statistical...

With the rapid development of economy and industrialization, environmental problems, especially haze pollution, are being more and more serious. When assessing the economic losses caused by haze, although the traditional quantitative method can show the amount of economic losses visually, there are also some inaccuracies in the calculation process....

With the rapid development of splicing manipulation, more and more negative effects have been brought. Therefore, the demand for image splicing detection algorithms is growing dramatically. In this paper, a new image splicing detection method is proposed which is based on convolutional neural network (CNN) with weight combination strategy. In the p...

Recently, hyperspectral image (HSI) classification methods based on convolutional neural networks (CNN) have developed rapidly with the advance of deep learning (DL) techniques. In order to more efficiently extract spatial and spectral features, we propose an end-to-end cubic convolutional neural network (Cubic-CNN) in this paper. The proposed Cubi...

Distinguishing malicious domain names generated by various domain generation algorithms (DGA) is critical for defending a network against sophisticated network attacks. In recent years, stealthy domain generation algorithms (SDGA) have been proposed and revealed significantly stronger stealthiness comparing to the traditional character-based DGA. E...

In this paper, a feature based on the Markov model in quaternion discrete cosine transform (QDCT) domain is proposed for double JPEG compression detection. Firstly, a given JPEG image is extracted from blocked images to obtain amplitude and three angles (, , and ). Secondly, when extracting the Markov features, we process the transition probability...

Due to the widespread use of the JPEG format, non-aligned double JPEG (NA-DJPEG) compression is very common in image tampering. Therefore, non-aligned double JPEG compression detection has attracted significant attention in digital forensics in recent years. In most of the previous detection algorithms, grayscale images are used directly, or color...

In order to improve the detection accuracy of hidden message in images, steganalysis features are selected as inputs for steganalysers. However, the existing Fisher criterion ignores the contribution of steganalysis feature components in dispersion to classification, which causes the useful feature components to be deleted, and decreases the detect...

Image splicing detection is of fundamental importance in digital forensics and therefore has attracted increasing attention recently. In this article, a color image splicing detection approach is proposed based on Markov transition probability of quaternion component separation in quaternion discrete cosine transform (QDCT) domain and quaternion wa...

This two-volume set LNCS 12239-12240 constitutes the refereed proceedings of the 6th International Conference on Artificial Intelligence and Security, ICAIS 2020, which was held in Hohhot, China, in July 2020. The conference was formerly called “International Conference on Cloud Computing and Security” with the acronym ICCCS. The total of 142 full...

This two-volume set LNCS 12239-12240 constitutes the refereed proceedings of the 6th International Conference on Artificial Intelligence and Security, ICAIS 2020, which was held in Hohhot, China, in July 2020. The conference was formerly called “International Conference on Cloud Computing and Security” with the acronym ICCCS. The total of 142 full...

The 3-volume set CCIS 1252 until CCIS 1254 constitutes the refereed proceedings of the 6th International Conference on Artificial Intelligence and Security, ICAIS 2020, which was held in Hohhot, China, in July 2020. The conference was formerly called “International Conference on Cloud Computing and Security” with the acronym ICCCS. The total of 178...

The 3-volume set CCIS 1252 until CCIS 1254 constitutes the refereed proceedings of the 6th International Conference on Artificial Intelligence and Security, ICAIS 2020, which was held in Hohhot, China, in July 2020. The conference was formerly called “International Conference on Cloud Computing and Security” with the acronym ICCCS. The total of 178...

The 3-volume set CCIS 1252 until CCIS 1254 constitutes the refereed proceedings of the 6th International Conference on Artificial Intelligence and Security, ICAIS 2020, which was held in Hohhot, China, in July 2020. The conference was formerly called “International Conference on Cloud Computing and Security” with the acronym ICCCS. The total of 178...

In order to extend the detection of JPEG compressed color images to solve the real-life problem, three-class classification forensics of JPEG compressed color images with the same quantization matrix is proposed. Since the previous methods treat detection of JPEG compressed color images as binary classification and JPEG compression with the same qu...

Splicing is a common tampering operation where the splicing area and the original image have different noise levels. In this paper, novel color spliced image localization based on quaternion principal component analysis (QPCA) and quaternion skewness is proposed. Firstly, the proposed algorithm uses QPCA to estimate the block-wise noise level. Then...

Detection of double Joint Photographic Experts Group (JPEG) compression is an important part of image forensics. Although methods in the past studies have been presented for detecting the double JPEG compression with a different quantization matrix, detection of double JPEG compression with the same quantization matrix is still a challenging proble...

Dynamic multi-watermarking embeds watermarks into the same region of a cover imperceptibly by simultaneously using different embedding rules. It is regarded as a solution to protect the cover’s copyright or track illegal distributors. Till now, few works have been done on multi-watermarking, especially dynamic multi-watermarking. In this paper, a n...

CAPTCHA (Completely Automated Public Turing test to tell Computers and Humans Apart) recognition is one of the most important branch in CAPTCHA research. The existing CAPTCHA recognition methods based on DCNN (Deep Convolutional Neural Network) have low recognition accuracy in confusion class. To solve this problem, a novel method of selective lear...

In the paper, we present a method estimating the linear Gaussian filter kernel by means of photo-response non-uniformity (PRNU) noise. In the past decade, a large number of works have proved that PRNU noise is a reliable fingerprint for digital image. Accordingly, PRNU noise has been widely used in such fields as camera identification and image man...

The convolutional neural network is widely popular for solving the problems of color image feature extraction. However, in the general network, the interrelationship of the color image channels is neglected. Therefore, a novel quaternion convolutional neural network (QCNN) is proposed in this paper, which always treats color triples as a whole to a...

This article proposes a novel robust reversible watermarking algorithm. The proposed watermarking scheme is reversible because the original image can be recovered after extracting watermarks from the watermarked image, as long as it is not processed by an attacker. The scheme is robust because watermarks can still be extracted from watermarked imag...

With the continuous development of cloud computing and big data technology, the use of cloud storage is more and more extensive, and a large amount of data is outsourced for public cloud servers, and the security problems that follow are gradually emerging. It can not only protect the data privacy of users, but also realize efficient retrieval and...

In the paper, a convolutional neural network based on quaternion transformation is proposed to detect median filtering for color images. Compared with conventional convolutional neural network, color images can be processed in a holistic manner in the proposed scheme, which makes full use of the correlation between RGB channels. And due to the use...

The healthcare data analytics is a demanding task at present due to enormous amount and diversity of information. Several algorithms were developed to analyze healthcare data with the objective to develop a non-invasive, unbiased and robust prediction system. The present study proposed a hybrid algorithm for healthcare data mining by using the Deci...

In this paper, a novel forensics scheme for color image is proposed in color quaternion wavelet transform (CQWT) domain. Compared to discrete wavelet transform (DWT), contourlet wavelet transform (CWT) and local binary patterns (LBP), CQWT processes a color image as a unit and so it can provide more forensics information to identify the photograph...

As for the fractional transforms, to date only fractional Fourier transform (FrFT) has been applicable to color images in a holistic manner by using the quaternion algebra, yet, discrete fractional random transform (DFRNT) with the useful intrinsic randomness is still to be explored. This paper first defines quaternion DFRNT (QDFRNT) which generali...

In this paper, an effective method based on the color quaternion wavelet transform (CQWT) for image forensics is proposed. Compared to discrete wavelet transform (DWT), the CQWT provides more information, such as the quaternion’s magnitude and phase measures, to discriminate between computer generated (CG) and photographic (PG) images. Meanwhile, w...

Estimating variances in noise is of key importance in many image processing applications, such as filtering, enhancement, quality assessment, and detecting forgery. For the existing detection methods that are based on inconsistencies in noise, the conventional approach is to estimate the noise variance of each region first and then identify the reg...

Matrix embedding (ME) can be exploited to improve the embedding efficiency for steganography by making fewer changes to the cover data. In ME, the sender and recipient agree on a parity check matrix (PCM) in advance. The PCM will be used by the sender to embed secret message into the cover data and later by the decoder to extract the embedded messa...

In this paper, a novel set of features based on Quaternion Wavelet Transform (QWT) is proposed for digital image forensics. Compared with DiscreteWavelet Transform (DWT) and ContourletWavelet Transform (CWT), QWT produces the parameters, i.e., one magnitude and three angles, which provide more valuable information to distinguish photographic (PG) i...

In this paper, we propose a new fragile watermarking scheme with high-quality recovery capability based on overlapping embedding strategy. The block-wise mechanism for tampering localization and the pixel-wise mechanism for content recovery are collaborated in the proposed scheme. With the assist of interleaving operation, reference bits are derive...

Image splicing detection is of fundamental importance in digital forensics and therefore has attracted increasing attention recently. In this paper, a color image splicing detection approach is proposed based on Markov transition probability of quaternion component separation in quaternion discrete cosine transform (QDCT) domain and quaternion wave...

Image splicing detection is of fundamental importance in digital forensics and therefore has attracted increasing attention recently. In this paper, a color image splicing detection approach is proposed based on Markov transition probability of quaternion component separation in quaternion discrete cosine transform (QDCT) domain and quaternion wave...

Recently, a reversible data hiding (RDH) method based on difference-pair-mapping (DPM) is proposed for better employing the image redundancy. The DPM is a specifically designed injective mapping defined on difference-pairs, and it is an extension of the conventional expansion and shifting techniques used in prior histogram-based RDH. However, any i...

This paper proposes a novel robust reversible data hiding scheme. This scheme is realized by two stages, which embed robust watermarks and fragile ones respectively. Firstly, the robust watermarks are embedded into the mid-low frequency coefficients after the image transformed from spatial domain to frequency domain. Then, side information that use...

In a spliced image, areas from different origins contain different noise features, which may be exploited as evidence for forgery detection. In this paper, we propose a noise level evaluation method for digital photos, and use the method to detect image splicing. Unlike most noise-based forensic techniques in which an AWGN model is assumed, the noi...

Multiwatermarking embeds multiple watermarks into the media content imperceptibly. It is regarded as the potential means to protect the media contents’ copyright and/or trace illegal redistributors in multi-user environments, e.g., collaborative media content production or layered media content distribution. However, there are still some open issue...

In this paper, we propose a new self-embedding watermarking scheme with hierarchical recovery capability. The binary bits in the adopted MSB layers are scrambled and individually interleaved with different extension ratios according to their importance to image visual quality. The interleaved data, which are regarded as reference bits for tampering...

In order to protect data privacy, image with sensitive or private information needs to be encrypted before being outsourced to a cloud service provider. However, this causes difficulties in image retrieval and data management. A privacy-preserving content-based image retrieval method based on orthogonal decomposition is proposed in the paper. The i...

In this paper, we propose a self-embedding fragile watermarking scheme using vector quantization (VQ) and index sharing. First, the principle contents of original image are compactly represented by a series of VQ indices. Then, after permutation, the binary bits of VQ indices are extended to generate reference-bits by a random binary matrix, which...

This paper presents a novel steganalysis scheme for detection of AMR (Adaptive Multi-Rate) audio steganography. AMR audio codec is used widely in mobile communication and mobile Internet system. Due to the modifiability of pitch delay, several AMR steganography schemes based on pitch delay modulation have emerged gradually and have high capacity an...

It is often important for reversible data hiding (RDH) to reduce the amount of image modifications for a given capacity. To this end, recently, a pairwise prediction-error expansion (pairwise PEE) is proposed to better exploit the image redundancy in the two-dimensional (2D) space. However, in conventional pairwise PEE, a drawback is that the pixel...

Some existing color image adaptive watermarking algorithms do not fully utilize the color information in the adaptive process, or do not consider the holistic property of the components of a color host image. To overcome these drawbacks, this paper proposes a color image adaptive watermarking algorithm based on Fractional Quaternion Fourier Transfo...

Given a natural image from the single sensor, the key task is to properly reconstruct the full color image. This paper presents an effectively demosaicking algorithm based on fully directional estimation using Bayer color filter array pattern. The proposed method smoothly keeps access to current reconstruction implementations, and outperforms the h...

Inter-frame forgery is a common type of video forgery in digital videos. In this paper, a method based on the consistency of quotient of mean structural similarity (QoMSSIM) is proposed. For original videos, the QoMSSIM are consistent, but in forgeries the consistency will be destroyed. First, the mean structural similarity (MSSIM) between every tw...

Pixel-value-ordering (PVO) technique refers to the process of first ranking the pixels in a block and then modifying the maximum/minimum for reversible data hiding (RDH). This paper discusses the PVO embedding in two-dimensional (2D) space and utilizes the prediction-error pair within a block for data embedding. We focus on not only the exploitatio...