Manifold-ranking embedded order preserving hashing for image semantic retrieval

Due to the storage and computational efficiency of hashing technology, it has proven a valuable tool for large scale similarity search. In many cases, the large scale data in real-world lie near some (unknown) low-dimensional and non-linear manifold. Moreover, Manifold Ranking approach can preserve the global topological structure of the data set more effectively than Euclidean Distance-based Ranking approach, which fails to preserve the semantic relevance degree. However, most existing hashing methods ignore the global topological structure of the data set. The key issue is how to incorporate the global topological structure of data set into learning effective hashing function. In this paper, we propose a novel unsupervised hashing approach, namely Manifold-Ranking Embedded Order Preserving Hashing (MREOPH). A manifold ranking loss is introduced to solve the issue of global topological structure preserving. An order preserving loss is introduced to ensure the consistency between manifold ranking and hamming ranking. A hypercubic quantization loss is introduced to learn discrete binary codes. The information theoretic regularization term is taken into consideration for preserving desirable properties of hash codes. Finally, we integrate them in a joint optimization framework for minimizing the information loss in each processing. Experimental results on three datasets for semantic search clearly demonstrate the effectiveness of the proposed method.     

Discriminative feature mining hashing for fine-grained image retrieval

With the development of multimedia technology, fine-grained image retrieval has gradually become a new hot topic in computer vision, while its accuracy and speed are limited due to the low discriminative high-dimensional real-valued embedding. To solve this problem, we propose an end-to-end framework named DFMH (Discriminative Feature Mining Hashing), which consists of the DFEM (Discriminative Feature Extracting Module) and SHCM (Semantic Hash Coding Module). Specifically, DFEM explores more discriminative local regions by attention drop and obtains finer local feature expression by attention re-sample. SHCM generates high-quality hash codes by combining the quantization loss and bit balance loss. Validated by extensive experiments and ablation studies, our method consistently outperforms both the state-of-the-art generic retrieval methods as well as fine-grained retrieval methods on three datasets, including CUB Birds, Stanford Dogs and Stanford Cars.     

JPEG encryption for image rescaling in the encrypted domain

This work proposes a novel protocol of encrypting the JPEG image suitable for image rescaling in the encrypted domain. To protect the privacy of original content, the image owner perturbs the texture and randomizes the structure of the JPEG image by enciphering the quantized Discrete Cosine Transform (DCT) coefficients. After receiving the encrypted JPEG image, the service provider generates a rescaled JPEG image by down-sampling the encrypted DCT coefficients. On the recipient side, the encrypted JPEG image rescaled by the service provider can be decrypted to a plaintext image with a lower resolution with the aid of encryption keys. Experimental results show that the proposed method has a good capability of rescaling the privacy-protected JPEG file.     

Robust and reversible image watermarking in homomorphic encrypted domain

This paper proposes a robust and reversible watermarking scheme for the encrypted image by using Paillier cryptosystem. In the proposed method, the original image is divided into a number of non-overlapping blocks sized by 8 × 8 and Paillier cryptosystem is applied to encrypt the pixels in each block. Firstly, a data hider can calculate the statistical values of encrypted blocks by employing modular multiplicative inverse (MMI) method and looking for a mapping table. Then a watermark sequence can be embedded into the encrypted image by shifting the histogram of the statistical values. On the receiver side, the shifted histogram can be obtained from both the encrypted image and the decrypted image. Furthermore, the embedded watermark can be extracted from the shifted histogram. The encrypted original image can be restored by employing inverse operations of histogram shifting. This is followed by a decryption operation to restore the original image. In the proposed method, the hidden bits can still be extracted correctly under some typical content-preserving operations, such as JPEG/JPEG2000 compression and additive Gaussian noise. Compared with the previous reversible watermarking methods in plaintext domain, the proposed method has satisfactory performance in image quality and robustness. Experimental results have shown the validity of the proposed method.     

Deep supervised hashing using quadratic spherical mutual information for efficient image retrieval

Several deep supervised hashing techniques have been proposed to allow for extracting compact and efficient neural network representations for various tasks. However, many deep supervised hashing techniques ignore several information-theoretic aspects of the process of information retrieval, often leading to sub-optimal results. In this paper, we propose an efficient deep supervised hashing algorithm that optimizes the learned compact codes using an information-theoretic measure, the Quadratic Mutual Information (QMI). The proposed method is adapted to the needs of efficient image hashing and information retrieval leading to a novel information-theoretic measure, the Quadratic Spherical Mutual Information (QSMI). Apart from demonstrating the effectiveness of the proposed method under different scenarios and outperforming existing state-of-the-art image hashing techniques, this paper provides a structured way to model the process of information retrieval and develop novel methods adapted to the needs of different applications.     

一种分形域基于内容的图像检索方法

基于内容的图像检索是多媒体、网络通信及计算机等应用研究领域的一项关键技术。该文提出了一种在分形压缩域直接进行基于内容的图像检索方法。该方法不需要对查询图像进行分形变换，因此可以提高检索速度，降低检索复杂度。仿真结果表明，使用该文提出的方法，能够有效地进行分形域基于内容的图像检索，比较大幅度地降低了检索时间，优于试验中其他3种方法。     

Secure and privacy‐preserving keyword search retrieval over hashed encrypted cloud data

Cloud computing (CC) is the universal area in which the data owners will contract out their pertinent data to the untrusted public cloud that permits the data users to retrieve the data with complete integrity. To give data privacy along with integrity, majority of the research works were concentrated on single data owner for secure searching of encrypted data via the cloud. Also, searchable encryption supports data user to retrieve the particular encrypted document from encrypted cloud data via keyword search (KS). However, these researches are not efficient for keyword search retrieval. To trounce such drawbacks, this paper proposes efficient secure and privacy‐preserving keyword search retrieval (SPKSR) system, in which the user retrieves the hashed encrypted documents over hashed encrypted cloud data. The proposed system includes three entities explicitly, (a) data owner (DO), (b) cloud server (CS), and (c) data users (DU). The owner outsources hashed encrypted documents set, along with generated searchable index tree to the CS. The CS hoards the hashed encrypted document collection and index tree structure. DU performs the “search” over the hashed encrypted data. Experimental results of the proposed system are analyzed and contrasted with the other existent system to show the dominance of the proposed system.     

A novel unsupervised multiple feature hashing for image retrieval and indexing (MFHIRI)

Recently, techniques that can automatically figure out the incisive information from gigantic visual databases are urging popularity. The existing multi-feature hashing method has achieved good results by fusing multiple features, but in processing these multi-features, fusing multi-features into one feature will cause the feature dimension to be very high, increasing the amount of calculation. On the one hand, it is not easy to discover the internal ties between different features. This paper proposes a novel unsupervised multiple feature hashing for image retrieval and indexing (MFHIRI) method to learn multiple views in a composite manner. The proposed scheme learns the binary codes of various information sources in a composite manner, and our scheme relies on weighted multiple information sources and improved KNN concept. In particular, here we adopt an adaptive weighing scheme to preserve the similarity and consistency among binary codes. Precisely, we follow the graph modeling theory to construct improved KNN concept, which further helps preserve different statistical properties of individual sources. The important aspect of improved KNN scheme is that we can find the neighbors of a data point by searching its neighbors’ neighbors. During optimization, the sub-problems are solved in parallel which efficiently lowers down the computation cost. The proposed approach shows consistent performance over state-of-the-art (three single-view and eight multi-view approaches) on three broadly followed datasets viz. CIFAR-10, NUS-WIDE and Caltech-256.     

SecureDL: A privacy preserving deep learning model for image recognition over cloud

The key benefits of cloud services such as low cost, access flexibility, and mobility have attracted worldwide users to utilize deep learning algorithms for computer vision. These cloud servers are maintained by third parties, where users are always concerned about sharing their confidential data with them. In this paper, we addressed these concerns for by developing SecureDL, a privacy-preserving image recognition model for encrypted data over cloud. The proposed block-based image encryption scheme is well designed to protect image’s visual information. The scheme constitutes an order-preserving permutation ordered binary number system and pseudo-random matrices. The proposed method is proved to be secure in a probabilistic viewpoint, and using various cryptographic attacks. Experiments are conducted over several image recognition datasets, and the trade-off analytics between the achieved recognition accuracy and data encryption is well described. SecureDL overcomes the storage and computational overheads that occur with fully-homomorphic and multi-party computation based secure recognition schemes.     

2DInpaint: A novel privacy-preserving scheme for image inpainting in an encrypted domain over the cloud

The low cost, agility, and mobility of cloud services for processing and storage data have attracted user’s attention today. Untrusted third parties support these services, and users are always concerned about utilizing them for personal data. Addressing these data-privacy issues for image inpainting over the cloud infrastructure(s), we propose a model, 2DInpaint, to perform image inpainting by protecting image information from an eavesdropping adversary. Inpainting is a technique for modifying an image in an undetectable manner with applications ranging from restoration of damaged photographs to object-removal and replacement of lost blocks in image coding and transmission. It can be accomplished by propagating the information in the isophotes direction of the desired region(s) from the neighborhood. Performing this propagation when the image is in the encrypted domain (ED) is a challenging dilemma. The challenge is addressed by employing a modified version of 2D-bicubic interpolation over the region to be inpainted in ED. The ramp secret sharing scheme is utilized to secure image information and to reduce storage overhead over the cloud server. 2DInpaint is proved to be information-theoretical secure in a probabilistic viewpoint and through various cryptographic attacks. The qualitative and quantitative results of 2DInpaint are analyzed under the scenarios of classical image inpainting, object-removal, and text-removal, and compared with the schemes in the plain domain. Moreover, no limitations related to the topology of the region to be inpainted are required using our approach. To the best of our knowledge, 2DInpaint is the first move towards image inpainting in the ED.     

Spline regression hashing for fast image search

Techniques for fast image retrieval over large databases have attracted considerable attention due to the rapid growth of web images. One promising way to accelerate image search is to use hashing technologies, which represent images by compact binary codewords. In this way, the similarity between images can be efficiently measured in terms of the Hamming distance between their corresponding binary codes. Although plenty of methods on generating hash codes have been proposed in recent years, there are still two key points that needed to be improved: 1) how to precisely preserve the similarity structure of the original data and 2) how to obtain the hash codes of the previously unseen data. In this paper, we propose our spline regression hashing method, in which both the local and global data similarity structures are exploited. To better capture the local manifold structure, we introduce splines developed in Sobolev space to find the local data mapping function. Furthermore, our framework simultaneously learns the hash codes of the training data and the hash function for the unseen data, which solves the out-of-sample problem. Extensive experiments conducted on real image datasets consisting of over one million images show that our proposed method outperforms the state-of-the-art techniques.     

ESOT: a new privacy model for preserving location privacy in Internet of Things

The Internet of Things (IoT) means connecting everything with every other thing through the Internet. In IoT, millions of devices communicate to exchange data and information with each other. During communication, security and privacy issues arise which need to be addressed. To protect information about users’ location, an efficient technique should be devised. Several techniques have already been proposed for preserving location privacy in IoT. However, the existing research lags in preserving location privacy in IoT and has highlighted several issues such as being specific or being restricted to a certain location. In this paper, we propose a new location privacy technique called the enhanced semantic obfuscation technique (ESOT) to preserve the location information of a user. Experimental results show that ESOT achieves improved location privacy and service utility when compared with a well-known existing approach, the semantic obfuscation technique.     

基于聚类的动态社交网络隐私保护方法

由于社交网络图结构的动态变化特性,需要采用有效的动态隐私保护方法。针对现有动态数据发布隐私保护方法中存在的攻击者背景知识单一、对图结构动态变化适应性较低等问题,提出基于聚类的动态图发布隐私保护方法。分析表明,该方法能抵御多种背景知识攻击,同时对社交网络图结构动态变化具有较好的适应性。     

End-to-end privacy preserving scheme for IoT-based healthcare systems

Preserving patients’ privacy is one of the most important challenges in IoT-based healthcare systems. Although patient privacy has been widely addressed in previous work, there is a lack of a comprehensive end-to-end approach that simultaneously preserves the location and data privacy of patients assuming that system entities are untrusted. Most of the past researches assume that parts of this end-to-end system are trustworthy while privacy may be threatened by insider attacks. In this paper, we propose an end-to-end privacy preserving scheme for the patients assuming that all main entities of the healthcare system (including sensors, gateways, and application providers) are untrusted. The proposed scheme preserves end-to-end privacy against insider threats as well as external attacks concerning the resource restrictions of the sensors. This scheme provides mutual authentication between main entities while preserves patients’ anonymity. Only the allowed users can access the real identity of patients alongside their locations and their healthcare information. Informal security analysis and formal security verification of the proposed protocol in AVISPA show that it is secure against impersonation, replay, modification, and man-in-the-middle attacks. Moreover, performance assessments show that the proposed protocol provides more security services without considerable growth in the computation overhead of the sensors. Also, it is shown that the proposed protocol diminishes the signaling overhead of the sensors and so their energy consumption compared to the literature at the expense of adding a little more signaling overhead to the gateways.

     

支持数据隐私保护的恶意加密流量检测确认方法

为解决基于机器学习的恶意加密流量检测易产生大量误报的问题,利用安全两方计算,在不泄露具体数据内容的前提下实现网络流量内容和入侵检测特征间的字符段比对.基于字符段比对结果,设计入侵检测特征匹配方法,完成关键词的精准匹配.为保证所提方法的有效执行,提出用户终端输入随机验证策略,使恶意用户终端难以使用任意数据参与安全两方计算...     

Using trusted computing for privacy preserving keystroke-based authentication in smartphones

Smartphones are increasingly being used to store personal information as well as to access sensitive data from the Internet and the cloud. Establishment of the identity of a user requesting information from smartphones is a prerequisite for secure systems in such scenarios. In the past, keystroke-based user identification has been successfully deployed on production-level mobile devices to mitigate the risks associated with naïve username/password based authentication. However, these approaches have two major limitations: they are not applicable to services where authentication occurs outside the domain of the mobile device—such as web-based services; and they often overly tax the limited computational capabilities of mobile devices. In this paper, we propose a protocol for keystroke dynamics analysis which allows web-based applications to make use of remote attestation and delegated keystroke analysis. The end result is an efficient keystroke-based user identification mechanism that strengthens traditional password protected services while mitigating the risks of user profiling by collaborating malicious web services. We present a prototype implementation of our protocol using the popular Android operating system for smartphones.     

Perceptual hashing for image authentication: A survey

Perceptual hashing is used for multimedia content identification and authentication through perception digests based on the understanding of multimedia content. This paper presents a literature review of image hashing for image authentication in the last decade. The objective of this paper is to provide a comprehensive survey and to highlight the pros and cons of existing state-of-the-art techniques. In this article, the general structure and classifications of image hashing based tamper detection techniques with their properties are exploited. Furthermore, the evaluation datasets and different performance metrics are also discussed. The paper concludes with recommendations and good practices drawn from the reviewed techniques.     

网络中多敏感属性数据发布隐私保护研究

随着网络技术的快速发展,许多社会网站被创建和使用,使得关于个人的社会网络信息大量被收集和发布.为了保证个人隐私的安全,本文提出了一个新的集值属性(k,l)anonymity隐私原则,开发了一个满足这个隐私原则的隐私算法来高效地处理集值属性数据流.并通过实验进行了验证本算法的高效率和有效性.