Abstract:Most existing image forensics approaches can effectively reveal image forgery by single image manipulation. However, two or more image manipulations, which are referred to be image operation chains, are usually involved in actual image tampering. In this paper, an image forensics approach is proposed to identify an image operation chain, which is possibly made up of three typical image manipulations including image sharping, scaling and median filtering. Since different image manipulations leave their distinct fingerprints, there also exist specific traces formed by interaction or superposition of different fingerprints left by multiple image manipulations in a specific order. Inspired by the rich image steganalytic features, we attempt to identify image operation chain by combining residue-based features in spatial domain and DCT domain. Experimental results show that the proposed approach achieves good detection result, which effectively discriminate multiple operation chains among different combinations of sharpening, scaling and median filtering. Moreover, the proposed method has good robustness against JPEG compression.

Alert Name:
Alerting to:
Authentication email will be sent to your email address in 24 hours
Frequency:
Email Message Format:	Plain text Graphical(HTML)

Complete the form below and we will recommend the selected titles to your friends on your behalf. * Indicates a required field.
Your name*:
Your email address*:
Recipient's name*:
Recipient's email address*:
(multiple recipient's names and email addresses should be separated with semicolons)
Your comments:	I thought you would find the page(s) useful.

Your name:
Your email address:
Recipient's name:
Recipient's email address:
(multiple recipient's names and email addresses should be separated with semicolons)
Your comments:	I thought you would find this page useful.

Disclaimer: This message was sent to your friend using the "Send it to a friend" facility on the Sciencepaper Online’ WWW site, http://www.paper.edu.cn/en. The Sciencepaper Online is not responsible for the content of this email, and anything said in this email does not necessarily reflect the Sciencepaper Online's views.


	1. Detection of image operation chains by residual-based hybrid features
	HU Lipin,YANG Gaobo
	Information Science and System Science 06 May 2020
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract:Most existing image forensics approaches can effectively reveal image forgery by single image manipulation. However, two or more image manipulations, which are referred to be image operation chains, are usually involved in actual image tampering. In this paper, an image forensics approach is proposed to identify an image operation chain, which is possibly made up of three typical image manipulations including image sharping, scaling and median filtering. Since different image manipulations leave their distinct fingerprints, there also exist specific traces formed by interaction or superposition of different fingerprints left by multiple image manipulations in a specific order. Inspired by the rich image steganalytic features, we attempt to identify image operation chain by combining residue-based features in spatial domain and DCT domain. Experimental results show that the proposed approach achieves good detection result, which effectively discriminate multiple operation chains among different combinations of sharpening, scaling and median filtering. Moreover, the proposed method has good robustness against JPEG compression.
	TO cite this article:HU Lipin,YANG Gaobo. Detection of image operation chains by residual-based hybrid features[OL].[ 6 May 2020] http://en.paper.edu.cn/en_releasepaper/content/4751889


	2. Statistical Modeling for Multiple Modes Facial Images using GND-PCA
	Qiao Xu
	Information Science and System Science 04 May 2017
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract: In this paper, we proposed a new approach called generalized N-dimensional principal component analysis (GND-PCA) for statistical appearance modeling of facial images with multiple modes including different people, different pose and different illumination. The facial images with multiple modes can be considered as high-dimensional data. GND-PCA can be used to treat the high-order dimensional data as a series of high-order tensors and calculate the bases on each mode-subspace in order to approximate the tensor accurately. GND-PCA can represent the high-order dimensional data of image ensembles more efficiently compared to the recently proposed ND-PCA method. MaVIC Database (KAO-Ritsumeikan Multi-angle View, Illumination and Cosmetic Facial Database) is used in our experiments and the results are compared with those obtained by conventional PCA and ND-PCA.
	TO cite this article:Qiao Xu. Statistical Modeling for Multiple Modes Facial Images using GND-PCA[OL].[ 4 May 2017] http://en.paper.edu.cn/en_releasepaper/content/4732187

Saved Papers

Saved Papers


	3. Pruned Convolutional Neural Network with two supervisory signals for Face Recognition
	ZOU Ying,LIU Xiaohong
	Information Science and System Science 11 November 2016
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract:Deep convolution neural network (CNN) has achieved a great success on face recognition techniques. But most of CNN models tend to be much deeper, which are at the expenses of high consumption of computation and storage. So, it is hard for these deep CNNs applied to mobile equipments because of poor computational and memory resources. To alleviate this issue, this paper optimizes a lightened baseline CNN model by adopting an additional contrastive loss to extract more discriminative features. To further reduce the number of parameters, a pruning strategy is tried to compress our model, which slightly improves accuracy on the LFW dataset with the compression ratio of 0.7. Finally, experimental results show that the proposed method achieve state-of-the-art results with much smaller size and fewer training data.
	TO cite this article:ZOU Ying,LIU Xiaohong. Pruned Convolutional Neural Network with two supervisory signals for Face Recognition[OL].[11 November 2016] http://en.paper.edu.cn/en_releasepaper/content/4709529


	4. An Element-based Model for Multi-document Summarization
	RONG Nan
	Information Science and System Science 21 March 2016
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract:The increasing availability of online information has necessitated intensive research in the area of automatic text summarization. In this paper, we propose an element-based model for multi-document summarization. First, we propose to model sentences from perspective of five major elements. Second, we cluster words according to their similarity computed through Word Affinity Force model. Third we propose a novelty detection algorithm to select sentences for summary. The experiment on the real-world data shows that the proposed model can find the key points of a document and generate fluent sentences.
	TO cite this article:RONG Nan. An Element-based Model for Multi-document Summarization[OL].[21 March 2016] http://en.paper.edu.cn/en_releasepaper/content/4681685


	5. A vehicular ad hoc networks routing algorithm based on clustering
	Tao Yang,TaoLing,Liu Jing
	Information Science and System Science 09 June 2015
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract:The vehicular ad hoc networks(VANETs) presents the uneven node density, fast moving speed,dynamic topology change etc.In order to improve the reliability of VANETs in information broadcasting, and reduce the redundancy of multi-hop broadcast, a protocol based on the clustering position-competition of vehicular ad hoc networks is proposed. Furthermore the concept of connectivity-stability is introduced.In the paper, The road is divided into segment model and intersection model.The cluster head and the optimal nodes in the cluster are chose as a radio relay nodes, which reduces the broadcast redundancy and improves the broadcast efficiency.At the same time, for the isolated nodes in the network ,carry-forwarding method is taken ,which improves the broadcast reliability in sparse areas and in "hole" areas.Finally, we use the NS2 simulation tool to.The simulation results show that the proposed routing protocol has a better comprehensive performance in delivery success rate, average broadcast time delay and the broadcast overhead.
	TO cite this article:Tao Yang,TaoLing,Liu Jing. A vehicular ad hoc networks routing algorithm based on clustering[OL].[ 9 June 2015] http://en.paper.edu.cn/en_releasepaper/content/4646209


	6. Load distribution and topological properties in urban road networks with metric spaces
	ZHENG Jianfeng
	Information Science and System Science 14 March 2015
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract:By considering the length and "width" of the road in urban transportation systems, this paper empirically investigates the topological structure properties for three realistic urban road networks in the United States, where road segments and junctions are represented by links and nodes, respectively. Links and nodes are embedded in metric spaces and the small-world effect is discussed in regular networks with metric spaces. The "width" of a road can be alternatively described by its capacity. Distributions of degree, capacity and length for these three realistic urban road networks are mainly analyzed. Finally, load distribution in these urban road networks is also explored.
	TO cite this article:ZHENG Jianfeng. Load distribution and topological properties in urban road networks with metric spaces[OL].[14 March 2015] http://en.paper.edu.cn/en_releasepaper/content/4633114


	7. Be Your Own Coach: Sport Motion Sensing using Smart Phone
	Qian Yichao,Li Shufang
	Information Science and System Science 31 December 2014
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract:Sports are the ways that people need to take exercises. However, better performance can not be achieved effectively just by exercises, highly cost coach and professional apparatus are probably needed, those are what common people can not afford. In this paper, we proposed a very low-cost alternative solution: monitoring body action during sports using smart phone to visualize the whole process of movement in fine time dimension. As a smart phone is cheap and public in current days, with our system, every common person can be a sport expert. We have implemented our system on real word smart phones with android system. As shown in this paper, take the 100 meter race for instance, our system can reproduce the velocity and acceleration during the race. Five phases of the race, Preparation, Start of Race, Acceleration, Sprint, Cross the End, can be automatically recognized and performance of athlete can be detailedly presented.
	TO cite this article:Qian Yichao,Li Shufang. Be Your Own Coach: Sport Motion Sensing using Smart Phone[OL].[31 December 2014] http://en.paper.edu.cn/en_releasepaper/content/4625662


	8. Application of MapReduce-Based Association Rule Data Mining on Operation of Telecommunication Network
	Miao Yu,Hou Chun-Ping
	Information Science and System Science 11 December 2014
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract:This paper based on the requirements and status of constructing data-sharing platform for telecom operators, proposed using the Hadoop open source framework for massive signaling data mining analysis to make up for the defects and deficiencies of traditional monitoring system. A Hadoop-based signaling mining platform and a MapReduce-based Apriori Algorithm is designed and implemented to discover the association rules on network service quality and data mutation. The experimental results show that, the method can effectively discover a variety of business factors and scenarios which lead to the mutation, and provide support for the upper application to solve the problem.
	TO cite this article:Miao Yu,Hou Chun-Ping. Application of MapReduce-Based Association Rule Data Mining on Operation of Telecommunication Network[OL].[11 December 2014] http://en.paper.edu.cn/en_releasepaper/content/4623060


	9. Robustness of centrality measures against network manipulation
	Qikai Niu,ZengAn,FanYing,Zengru Di
	Information Science and System Science 19 November 2014
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract:Node centrality is an important quantity for complex networks as it is related to many applications ranging from the prediction of network structure to the control of dynamics on networks. In the literature, much effort has been devoted to design new centrality measurements. However, the reliability of these centrality measurements haven't been fully understood. Many real networks are facing different kinds of manipulations such as addition, removal or rewiring of links. In this paper, we focus on the robustness of classic centrality measures against network manipulation. Our analysis is based on both artificial and real networks. We find that the centrality measurements are generally more robust in heterogenous networks. Moreover, the top part of the centrality ranking is more resistant to manipulation. Among the centrality measures we considered, the eigenvector centrality is in general the most robust one.?????
	TO cite this article:Qikai Niu,ZengAn,FanYing, et al. Robustness of centrality measures against network manipulation[OL].[19 November 2014] http://en.paper.edu.cn/en_releasepaper/content/4618979


	10. Person Re-identification with Data-Driven Features
	Xiang Li, Jinyu Gao, Xiaobin Chang, Yuting Mai, Wei-Shi Zheng
	Information Science and System Science 08 September 2014
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract:Human-specified appearance features are widely used for person re-identification at present, such as color and texture histograms. Often, thesefeatures are limited by the subjective appearance of pedestrians. This paper presents a new representation to re-identification that incorporates data-driven features to improve the reliability and robustness in person matching. Firstly, we utilize a deep learning network, namely PCA Network, to learn data-driven features from person images. The features mine more discriminative cues from pedestrian data and compensate the drawback of human-specified features. Then the data-driven features and common human-specified features are combined to produce a final representation of each image. The so-obtained enriched Data-driven Representation (eDR) has been validated through experiments on two person re-identification datasets, demonstrating that the proposed representation is effective for person matching. That is, the data-driven features facilitatemore accurate re-identification when they are fused together with the human-specified features.
	TO cite this article:Xiang Li, Jinyu Gao, Xiaobin Chang, et al. Person Re-identification with Data-Driven Features[OL].[ 8 September 2014] http://en.paper.edu.cn/en_releasepaper/content/4607551

	Check out RSS, or use RSS reader to subscribe this item