Search

Cross-modal retrieval is such a challenging topic that traditional global representations would fail to bridge the semantic gap between images and texts to a satisfactory level. Using local features from images and words from documents directly can be more robust for the scenario with large intraclass variations and small interclass discrepancies. In this paper, we propose a novel unsupervised binary coding algorithm called binary set embedding (BSE) to obtain meaningful hash codes for local features from the image domain and words from text domain. Understanding image features with the word vectors learned from the human language instead of the provided documents from data sets, BSE can map samples into a common Hamming space effectively and efficiently where each sample is represented by the sets of local feature descriptors from image and text domains. In particular, BSE explores relationship among local features in both feature level and image (text) level, which can balance the sensitivity of each other. Furthermore, a recursive orthogonalization procedure is applied to reduce the redundancy of codes. Extensive experiments demonstrate the superior performance of BSE compared with state-of-the-art cross-modal hashing methods using either image or text queries.

We derive bounds for the ball L p -discrepancies in the Hamming space for 0 p ∞ and p = ∞ . Sharp estimates of discrepancies have been obtained for many spaces such as the Euclidean spheres and more general compact Riemannian manifolds. In the present paper, we show that the behavior of discrepancies in the Hamming space differs fundamentally because the volume of the ball in this space depends on its radius exponentially while such a dependence for the Riemannian manifolds is polynomial.

The space-partitioning based hashing techniques are widely used to represent high-dimensional data points as bit-codes. Although Binary Search Trees (BSTs) can be used for storing bit-codes, their size grows exponentially with code length. In practice, such a tree turns out to be highly sparse, increasing the miss-rate of nearest neighbor searches. We present Compressed BST of Inverted hash tables (ComBI), a geometrically motivated compression technique for BSTs. ComBI enables fast and approximate nearest neighbor searches without a significant memory footprint over BSTs. We show that approximate search in ComBI can perform competitively to an exact search algorithm in retrieving the nearest neighbors in a Hamming space. On a database containing ∼80M samples, ComBI yields an average precision of 0.90, at ∼4X - ∼296X improvements in run-time across different code lengths when compared to MIH, a widely used exact search method. On a database consisting of 1B samples, this value of precision (0.90) is reached at ∼4X - ∼19X improvements in run-time.

Content-based image retrieval (CBIR) shows great significance in clinical decision-making, which explores the visual content of medical images rather than keywords, tags, or descriptions. It provides doctors an image-guided approach to explore relevant cases that could offer doctors instructive reference. Mammogram screening has been known to be widely used in the early stage diagnosis of breast cancer and could reduce its morbidity and mortality. In this paper, we aim to develop a scalable CBIR method for a large repository of mammogram. To this end, we extend the original Anchor Graph Hashing (AGH) and propose a new unsupervised hashing algorithm, named as composite AGH with iterative quantization (C-AGH-ITQ), which compresses mammographic regions of interest (ROIs) into compact binary codes and enables real-time searching in Hamming space. Multimodal features and different distance metrics are integrated, performing upon a composite Anchor Graph. To improve the effectiveness of the hash code, quantization error is further iteratively minimized by introducing an orthogonal rotation matrix. We evaluate the presented C-AGH-ITQ algorithm on a data set of 11 533 mammographic ROIs obtained from the Digital Database for Screening Mammography. Our method obtains more than 84% retrieval precision and 93% classification accuracy (using $k$ NN prediction), which demonstrates that hash codes produced by C-AGH-ITQ well capture the visual similarities between mammographic images. In addition, since C-AGH-ITQ ensures linear complexity of the training procedure and constant time for query, our system is readily applicable to large-scale mammogram databases and has the potential to provide abundant clinical cases as reference.

Binary code learning, also known as hashing, has received increasing attention in large-scale visual search. By transforming high-dimensional features to binary codes, the original Euclidean distance is approximated via Hamming distance. More recently, it is advocated that it is the manifold distance, rather than the Euclidean distance, that should be preserved in the Hamming space. However, it retains as an open problem to directly preserve the manifold structure by hashing. In particular, it first needs to build the local linear embedding in the original feature space, and then quantize such embedding to binary codes. Such a two-step coding is problematic and less optimized. Besides, the off-line learning is extremely time and memory consuming, which needs to calculate the similarity matrix of the original data. In this paper, we propose a novel hashing algorithm, termed discrete locality linear embedding hashing (DLLH), which well addresses the above challenges. The DLLH directly reconstructs the manifold structure in the Hamming space, which learns optimal hash codes to maintain the local linear relationship of data points. To learn discrete locally linear embedding codes, we further propose a discrete optimization algorithm with an iterative parameters updating scheme. Moreover, an anchor-based acceleration scheme, termed Anchor-DLLH , is further introduced, which approximates the large similarity matrix by the product of two low-rank matrices. Experimental results on three widely used benchmark data sets, i.e., CIFAR10, NUS-WIDE, and YouTube Face, have shown superior performance of the proposed DLLH over the state-of-the-art approaches.

With the rapid growth of the number of images on the Internet, it has become more necessary to ensure the content security of images. The key problem is retrieving relevant images from the large database. Binary embedding is an effective way to improve the efficiency of calculating similarities for image content security as binary code is storage efficient and fast to compute. It tries to convert real-valued signatures into binary codes while preserving similarity of the original data, and most binary embedding methods quantize each projected dimension to one bit (presented as 0/1). As a consequence, it greatly decreases the discriminability of original signatures. In this paper, we first propose a novel triple-bit quantization strategy to solve the problem by assigning 3-bit to each dimension. Then, asymmetric distance algorithm is applied to re-rank candidates obtained from Hamming space for the final nearest neighbors. For simplicity, we call the framework triple-bit quantization with asymmetric distance (TBAD). The inherence of TBAD is combining the best of binary codes and real-valued signatures to get nearest neighbors quickly and concisely. Moreover, TBAD is applicable to a wide variety of embedding techniques. Experimental comparisons on BIGANN set show that the proposed method can achieve remarkable improvement in query accuracy compared to original binary embedding methods.

Hashing based approximate nearest neighbor (ANN) search techniques have attracted considerable attention in media search community because of its good potential for low storage cost and fast query speed. Many hashing based ANN search methods have been proposed; but, most of them just consider to keep the similarity relationship of data points during mapping instead of topology of data. It is well known that Self-Organizing Map can keep topology structure while conducting mapping task. Motivated by this, in this paper, we propose a Self-Organizing Map based hashing framework–SOMH, which cannot only keep similarity relationship, but also preserve topology of data. Specifically, in SOMH, Self-Organizing Map is introduced to map data points into hamming space. In addition, in order to make it work well on short and long binary codes, we propose a relaxed version of SOMH and a product space SOMH, respectively. For the optimization problem of the relaxed SOMH, we also present an iterative solution. Moreover, we further propose an extended version of SOMH, which can work well on multimodal data search task, i.e., cross-modal search. To test the performance of these proposed algorithms, we conduct experiments on three data sets–SIFT1M, GIST1M and Wiki (a multimodal dataset). Experimental results show that SOMH can outperform or is comparable to several state-of-the-arts.

© 2017 IEEE. Hashing technique has become an effective method for information retrieval due to the fast calculation of the Hamming distance. However, with the continuous growth of data coming from the Internet, the online update of hashing on the massive social data becomes very time-consuming. To alleviate this issue, in this paper, we propose a novel updating technique for hashing methods, namely Hamming Subspace Learning (HSL). The motivation of HSL is to generate a low-dimensional Hamming subspace from a high-dimensional Hamming space by selecting representative hash functions. Through HSL, we aim to improve the speed of updating binary codes for all samples. We present a method for Hamming subspace learning based on greedy selection strategy and the Distribution Preserving Hamming Subspace learning (DHSL) algorithm by designing a novel loss function. The experimental results demonstrate that the HSL is effective to improve the speed of online updating and the performance of hashing algorithm.

There has been increasing interest in learning compact binary codes for large-scale image data representation and retrieval. In most existing hashing-based methods, high-dimensional vectors are hashed into Hamming space, and the similarity between two vectors is approximated by the Hamming distance between their binary codes. Although hashing-based binary codes generation methods were widely used, Product Quantization (PQ) has been shown to be more accurate than various hashing-based methods, largely due to its lower quantization distortions and more precise distance computation. However, it is still a challenging problem to generalize PQ to accommodate arbitrary kernels. In this paper, we demonstrate how to employ arbitrary kernel functions in a PQ scheme. First, we propose a Kernelized PQ (KPQ) method based on composite kernels, which serves as a basic framework by making the decomposition of implicit feature space possible. Furthermore, we propose a Kernelized Optimized PQ (KOPQ) method to generalize Optimized Product Quantization (OPQ) to an arbitrary implicit feature space. Finally, we propose a Supervised KPQ (SKPQ) to improve the performance of semantic neighbor search. Both methods are variations of KPQ with the incorporation of their corresponding core techniques, KPCA and KCCA respectively, to the basic KPQ framework. Experiments involving three notable datasets show that KPQ, KOPQ and SKPQ can outperform the state-of-the-art methods for a similarity search in feature space or semantic search.

In recent years, learning based hashing becomes an attractive technique in large-scale image retrieval due to its low storage and computation cost. Hashing methods map each high-dimensional vector onto a low-dimensional hamming space by projection operators. However, when processing high dimensional data retrieval, many existing methods including hashing cost a majority of time on projection operators. In this paper, we solve this problem by implementing a sparsity regularizer. On one hand, due to the sparse property of the projection matrix, our method effectively lower both the storage and computation cost. On the other hand, we reduce the effective number of parameters involved in the learned projection matrix according to sparsity regularizer, which helps avoid overfitting problem. Without relaxing binary constraints, an iterative scheme jointly optimizing the objective function directly was given, which helps to obtain effective and efficient binary codes. We evaluate our method on three databases and compare it with some state-of-the-art hashing methods. Experimental results demonstrate that our method outperforms the comparison approaches.

Existing hashing methods normally define certain specific forms of hash functions, after which an objective function can be formulated to optimize the loss on training set to learn the parameters. However, in this way, the hash function will be tightly coupled with the generated objective in most cases. Moreover, since the objectives are generally formulated with binary quantization, most of them are nonconvex, which makes the optimization difficult and consequently decreases the similarity preserving performance of hashing. To solve this problem, we propose a novel pairwise correlation preserving framework to learn compact binary codes for hashing. First, we project each data into a metric space and represent it as a vector encoding the underlying local and global structure by pairwise correlation learning. Afterwards, pairwise correlation reconstruction (PCR), is further proposed to preserve the correlations of data between the metric space and the hamming space to learn binary codes. The PCR model is convex. Moreover, no specific hash functions are needed to be predefined and the steps of correlation learning and reconstruction are independent. The above characteristics make the optimization of PCR easily and efficiently, and thus leads to better preservation of data similarity in hamming space.

This paper proposes a novel framework to help biologists explore and analyze neurons based on retrieval of data from neuron morphological databases. In recent years, the continuously expanding neuron databases provide a rich source of information to associate neuronal morphologies with their functional properties. We design a coarse-to-fine framework for efficient and effective data retrieval from large-scale neuron databases. In the coarse-level, for efficiency in large-scale, we employ a binary coding method to compress morphological features into binary codes of tens of bits. Short binary codes allow for real-time similarity searching in Hamming space. Because the neuron databases are continuously expanding, it is inefficient to re-train the binary coding model from scratch when adding new neurons. To solve this problem, we extend binary coding with online updating schemes, which only considers the newly added neurons and update the model on-the-fly, without accessing the whole neuron databases. In the fine-grained level, we introduce domain experts/users in the framework, which can give relevance feedback for the binary coding based retrieval results. This interactive strategy can improve the retrieval performance through re-ranking the above coarse results, where we design a new similarity measure and take the feedback into account. Our framework is validated on more than 17,000 neuron cells, showing promising retrieval accuracy and efficiency. Moreover, we demonstrate its use case in assisting biologists to identify and explore unknown neurons.

We develop and apply combinatorial algorithms for investigation of the feasible distance distributions of binary orthogonal arrays with respect to a point of the ambient binary Hamming space utilizing constraints imposed from the relations between the distance distributions of connected arrays. This turns out to be strong enough and we prove the nonexistence of binary orthogonal arrays of parameters (length, cardinality, strength) = ( 9 , 96 , 4 ) , ( 10 , 192 , 5 ) , ( 10 , 112 , 4 ) , ( 11 , 224 , 5 ) , ( 11 , 112 , 4 ) and ( 12 , 224 , 5 ) , resolving the first cases where the existence was undecided so far. For the existing arrays our approach allows substantial reduction of the number of feasible distance distributions which could be helpful for classification results (uniqueness, for example).

We revisit the classic DBSCAN algorithm by proposing a series of strategies to improve its robustness to various densities and its efficiency. Unlike the original DBSCAN, we first use the binary local sensitive hashing (LSH) which enables faster region query for the k neighbors of a data point. The binary data representation method based on k neighborhood is then proposed to map the dataset into the Hamming space for faster cluster expansion. We define a core point based on binary influence space to enhance the robustness to various densities. Also, we propose a seed point selection method, which is based on influence space and k neighborhood similarity, to select some seed points instead of all the neighborhood during cluster expansion. Consequently, the number of region queries can be decreased. The experimental results show that the improved algorithm can greatly improve the clustering speed under the premise of ensuring better algorithm clustering accuracy, especially for large-scale datasets.

Hash-based methods can achieve a fast similarity search by representing high-dimensional data with compact binary codes. However, the spatial structure in row images was always lost in most previous methods. In this paper, a novel Locally Linear Spatial Pyramid Hash(LLSPH) algorithm is developed for the task of fast image retrieval. Unlike the conventional approach, the spatial extent of image features is exploited in our method. The spatial pyramid structure is used both to construct binary hash codes and to increase the discriminability of the description. To generate interpretable binary codes, the proposed LLSPH method captures the spatial characteristics of the original SPM and generates a low-dimensional sparse representation using multi-dictionaries Locality-constrained Linear Coding(MD_LLC). LLSPH then converts the low-dimensional data into Hamming space by the TF-IDF binarization rule. Our experimental results show that our LLSPH method can outperform several state-of-the-art hashing algorithms on the Caltech256 and ImageNet-500 datasets.

In this work, a simple angular reconstruction based binary descriptor (ARBD) strategy is proposed for face recognition. Compared with previous learning-based hashing methods such as compact binary face descriptor (CBFD) and sparse projection matrix binary descriptor (SPMBD) which measure data similarity with the Euclidean distance, our ARBD focuses on using the cosine similarity to generate binary codes. Specifically, the angular reconstruction term is added to the objective function to minimize the reconstruction error. Furthermore, an efficient algorithm based on the augmented Lagrange method (ALM) is designed to explicitly address the discrete constraint in the Hamming space. We also perform pooling on the learned binary codes with the unsupervised clustering manner to improve their discriminative ability. The results on two face datasets (i.e., CAS-PEAL-R1 and PaSC) demonstrate the superior performance of our ARBD over other existing face recognition methods.

Knowledge graph (KG) embedding techniques represent entities and relations as low-dimensional, continuous vectors, and thus enables machine learning models to be easily adapted to KG completion and querying tasks. However, learned dense vectors are inefficient for large-scale similarity computations. Learning-to-hash is to learn compact binary codes from high-dimensional input data and provides a promising way to accelerate efficiency by measuring Hamming distance instead of Euclidean distance or dot-product. Unfortunately, most of learning-to-hash methods cannot be directly applied to KG structure encoding. In this paper, we introduce a novel framework for encoding incomplete KGs and graph queries in Hamming space. To preserve KG structure information from embeddings to hash codes and address the ill-posed gradient issue in optimization, we utilize a continuation method with convergence guarantees to jointly encode queries and KG entities with geometric operations. The hashed embedding of a query can be utilized to discover target answers from incomplete KGs whilst the efficiency has been greatly improved.We compared our model with state-of-the-art methods on real-world KGs. Experimental results show that our framework not only significantly speeds up the searching process, but also provides good results for unanswerable queries caused by incomplete information.

Most Cross-modal hashing methods do not sufficiently exploit the discrimination power of semantic information when learning hash codes, while often involving time-consuming training procedures for large-scale dataset. To tackle these issues, we first formulate the learning of similarity-preserving hash codes in terms of orthogonally rotating the semantic data to hamming space, and then propose a novel Fast Semantic Preserving Hashing (FSePH) approach to large-scale cross-modal retrieval. Specifically, FSePH introduces an orthonormal basis to regress the targeted hash codes of training examples to their corresponding reasonably relaxed class labels, featuring significantly reducing the quantization error. Meanwhile, an effective optimization algorithm is derived for modality-specific projection function learning and an efficient closed-form solution for hash code learning, which are computationally tractable. Extensive experiments have shown that the proposed FSePH approach runs sufficiently fast, and also significantly improves the retrieval performances over the state-of-the-arts.

Cross-modal hashing, due to its low storage cost and high query speed, has been successfully used for similarity search in multimedia retrieval applications. It projects high-dimensional data into a shared isomorphic Hamming space with similar binary codes for semantically-similar data. In some applications, all modalities may not be obtained or trained simultaneously for some reasons, such as privacy, secret, storage limitation, and computational resource limitation. However, most existing cross-modal hashing methods need all modalities to jointly learn the common Hamming space, thus hindering them from handling these problems. In this paper, we propose a novel approach called Separated Variational Hashing Networks (SVHNs) to overcome the above challenge. Firstly, it adopts a label network (LabNet) to exploit available and nonspecific label annotations to learn a latent common Hamming space by projecting each semantic label into a common binary representation. Then, each modality-specific network can separately map the samples of the corresponding modality into their binary semantic codes learned by LabNet. We achieve it by conducting variational inference to match the aggregated posterior of the hashing code of LabNet with an arbitrary prior distribution. The effectiveness and efficiency of our SVHNs are verified by extensive experiments carried out on four widely-used multimedia databases, in comparison with 11 state-of-the-art approaches.

Recent years have witnessed the success of learning to hash in fast large-scale image retrieval. As deep learning has shown its superior performance on many computer vision applications, recent designs of learning-based hashing models have been moving from shallow ones to deep architectures. However, based on our analysis, we find that gradient descent based algorithms used in deep hashing models would potentially cause hash codes of a pair of training instances to be updated towards the directions of each other simultaneously during optimization. In the worst case, the paired hash codes switch their directions after update, and consequently, their corresponding distance in the Hamming space remain unchanged. This makes the overall learning process highly inefficient. To address this issue, we propose a new deep hashing model integrated with a novel gradient attention mechanism. Extensive experimental results on three benchmark datasets show that our proposed algorithm is able to accelerate the learning process and obtain competitive retrieval performance compared with state-of-the-art deep hashing models.

In recent years, learning-based hashing has gained increasing research interest in face recognition due to its encouraging storage and computational efficiency. However, learning high-quality binary codes that yield appealing recognition performance is still a challenge. In addition, since the neighborhood structure is important for discrimination, it should be taken into account to capture useful nearest neighbors. In this paper, we propose a novel and efficient graph-based hashing model, called local feature hashing with graph regularized binary auto-encoder (LFH-GRBA), to learn feature representations for face recognition. It considers the semantic neighborhood of the face data and seeks to reconstruct a face image from the learned binary codes. Specifically, the graph Laplacian is incorporated into the binary auto-encoder as a regularizer to exploit the semantical neighborhood information of the face data. To make such a model computational efficiency, a tractable alternating optimization approach is proposed to solve the objective function, yielding high-quality binary codes to well capture the neighborhood structure and provide high discrimination. Moreover, the discrete cyclic coordinate descent (DCC) method is adopted to directly learn binary codes in the Hamming space to eliminate the accumulated quantization errors. Extensive experimental results on three widely used datasets including FERET, LFW and PaSC datasets demonstrate that the proposed LFH-GRBA model outperforms most state-of-the-art face representation methods.

Deep hashing enables computation and memory efficient image search through end-to-end learning of feature representations and binary codes. While linear scan over binary hash codes is more efficient than over the high-dimensional representations, its linear-time complexity is still unacceptable for very large databases. Hamming space retrieval enables constant-time search through hash lookups, where for each query, there is a Hamming ball centered at the query and the data points within the ball are returned as relevant. Since inside the Hamming ball implies retrievable while outside irretrievable, it is crucial to explicitly characterize the Hamming ball. The main idea of this work is to directly embody the Hamming radius into the loss functions, leading to Maximum-Margin Hamming Hashing (MMHH), a new model specifically optimized for Hamming space retrieval. We introduce a max-margin t-distribution loss, where the t-distribution concentrates more similar data points to be within the Hamming ball, and the margin characterizes the Hamming radius such that less penalization is applied to similar data points within the Hamming ball. The loss function also introduces robustness to data noise, where the similarity supervision may be inaccurate in practical problems. The model is trained end-to-end using a new semi-batch optimization algorithm tailored to extremely imbalanced data. Our method yields state-of-the-art results on four datasets and shows superior performance on noisy data.

В работе рассмотрены основные положения обратимости как новой парадигмы развития вычислительной техники. Первые разделы носят обзорный характер. Показана неизбежность т. н. теплового проклятия при сохранении традиционной парадигмы создания средств ВТ. Изложены основы обратимой логики, рассмотрены основные обратимые логические элементы и модели обратимых вычислений, в т. ч. обратимые клеточные автоматы. Кратко рассмотрены обратимые языки программирования. Во второй части затронуты основные вопросы логического синтеза схем из обратимых элементов и физическая реализация обратимой схемотехники. Кратко описана проблематика синтеза отказоустойчивых схем в парадигме обратимой схемотехники. Предлагается техника синтеза сбоеустойчивых обратимых элементов в хэмминговом пространстве и описываются некоторые такие схемы. Далее рассматривается проблематика применения схем из обратимых логических элементов в криптографии. Описывается предлагаемая общая схема создания обратимых схем с уборкой мусора , предназначенных для криптографических применений., The paper considers the main provisions of reversibility as a new paradigm for the development of computer technology. The first sections are of an overview nature. The inevitability of the so-called heat curse while maintaining the traditional paradigm of creating means of computer engineering. The fundamentals of reversible logic are presented, the main reversible logic elements and models of reversible computations, including reversible cellular automata, are considered. Reversible programming languages are briefly reviewed. The second part addresses the basic issues of the logical synthesis of circuits from reversible elements and the physical implementation of reversible circuitry. The synthesis of fault-tolerant circuits in the paradigm of reversible circuitry is briefly described. A technique for synthesizing fault-tolerant reversible elements in a hamming space is proposed and some such schemes are described. Next, the problems of using circuits of reversible logic elements in cryptography are considered. The proposed general scheme for creating reversible schemes with garbage collection intended for cryptographic applications is described., №3 (2020)

Cross-modal hashing, which searches nearest neighbors across different modalities in the Hamming space, has become a popular technique to overcome the storage and computation barrier in multimedia retrieval recently. Although dozens of cross-modal hashing algorithms are proposed to yield compact binary code representation, applying exhaustive search in a large-scale dataset is impractical for the real-time purpose, and the Hamming distance computation suffers inaccurate results. In this paper, we propose a novel index scheme over binary hash codes in cross-modal retrieval. The proposed indexing scheme exploits a few binary bits of the hash code as the index code. Based on the index code representation, we construct an inverted index structure to accelerate the retrieval efficiency and train a neural network to improve the indexing accuracy. Experiments are performed on two benchmark datasets for retrieval across image and text modalities, where hash codes are generated by three cross-modal hashing methods. Results show the proposed method effectively boosts the performance over the benchmark datasets and hash methods.

Venue discovery using real-world multimedia data has not been investigated thoroughly. We are referring to business and travel locations as venues in this study and aim to improve the efficiency of venue discovery by hashing. Most existing supervised cross-modal hashing methods map data in different modalities to Hamming space, where the semantic information is exploited to supervise data of different modalities during the training stage. However, previous works neglect pairwise similarity between data in different modalities, which lead to degraded performance of hashing function learning. To address this issue, we propose a supervised Generative Adversarial Cross-modal Hashing method by Transferring Pairwise Similarities (SGACH-TPS). This work has three significant contributions: i) we propose a model for making efficient venue discovery, ii) the supervised generative adversarial network can construct a hash function to map multimodal data to a common hamming space. iii) a simple transfer training strategy for the adversarial network is suggested to supervise data in different modalities where the pairwise similarity is transferred to the fine-tuning stage of training. Evaluation on the new WikiVenue dataset confirms the superiority of the proposed method.

Similarity search plays a major role in a large variety of database applications where high-dimensional data, usually derived from images and videos, are mapped onto indexed structures. Its essence relies on a k-nearest-neighbors (kNN) algorithm to find similar contents given a specific input data. An efficient way to store and speedup such searches is to apply kNN in Hamming space by encoding the data as binary and splitting it into multiple hash tables. This paper proposes an efficient Hardware-Accelerated Similarity Search co-processor architecture using Multi-Index Hashing as storage structure. The accelerator is specified in Verilog hardware description language and implemented in a Xilinx low-cost Zynq FPGA. Performance, circuit-area and power consumption results are presented, with the accelerator being up to 13x and 18x faster than the corresponding C/C++ code on the ARM host processor when running the SIFT and GIST datasets, respectively, while also requiring less power.

Existing data-dependent hashing methods usually learn hash functions from pairwise or triplet data relationships, which only capture the data similarity locally, and often suffer from low learning efficiency and low collision rate. In this work, we propose a new \emph{global} similarity metric, termed as \emph{central similarity}, with which the hash codes of similar data pairs are encouraged to approach a common center and those for dissimilar pairs to converge to different centers, to improve hash learning efficiency and retrieval accuracy. We principally formulate the computation of the proposed central similarity metric by introducing a new concept, i.e., \emph{hash center} that refers to a set of data points scattered in the Hamming space with a sufficient mutual distance between each other. We then provide an efficient method to construct well separated hash centers by leveraging the Hadamard matrix and Bernoulli distributions. Finally, we propose the Central Similarity Quantization (CSQ) that optimizes the central similarity between data points w.r.t.\ their hash centers instead of optimizing the local similarity. CSQ is generic and applicable to both image and video hashing scenarios. Extensive experiments on large-scale image and video retrieval tasks demonstrate that CSQ can generate cohesive hash codes for similar data pairs and dispersed hash codes for dissimilar pairs, achieving a noticeable boost in retrieval performance, i.e. 3\%-20\% in mAP over the previous state-of-the-arts. The code is at: \url{https://github.com/yuanli2333/Hadamard-Matrix-for-hashing}, CVPR2020, Codes: https://github.com/yuanli2333/Hadamard-Matrix-for-hashing

Nowadays, the network intrusion detection has been an important issue in IoT. Although machine learning based methods seem to be promising in traditional network intrusion detection, these methods can hardly meet some demands of IoT. For example, unknown classes of flows are produced frequently in IoT, leading to classifiers training repeatly. To address this issue, we proposed a network intrusion detection framework based on learning to hash in this paper, which can reduce computation overhead significantly while avoiding frequent training of classifiers. The proposed framework consists of a hashing encoding module and an anomaly detection module with optimized k-NN classifier based on data distribution ratio. Moreover, the multi-index hashing is applied for fast and accurate search in Hamming space. Experimental results show that the proposed framework can detect various attacks and outperform the traditional intrusion detector.

Cross-modal hashing has attracted considerable attention for large-scale multimodal retrieval task. A majority of hashing methods have been proposed for cross-modal retrieval. However, these methods inadequately focus on feature learning process and cannot fully preserve higher-ranking correlation of various item pairs as well as the multi-label semantics of each item, so that the quality of binary codes may be downgraded. To tackle these problems, in this paper, we propose a novel deep cross-modal hashing method, called Adversary Guided Asymmetric Hashing (AGAH). Specifically, it employs an adversarial learning guided multi-label attention module to enhance the feature learning part which can learn discriminative feature representations and keep the cross-modal invariability. Furthermore, in order to generate hash codes which can fully preserve the multi-label semantics of all items, we propose an asymmetric hashing method which utilizes a multi-label binary code map that can equip the hash codes with multi-label semantic information. In addition, to ensure higher-ranking correlation of all similar item pairs than those of dissimilar ones, we adopt a new triplet-margin constraint and a cosine quantization technique for Hamming space similarity preservation. Extensive empirical studies show that AGAH outperforms several state-of-the-art methods for cross-modal retrieval.

© 1992-2012 IEEE. In this paper, we study the problem of cross-modal retrieval by hashing-based approximate nearest neighbor search techniques. Most existing cross-modal hashing works mainly address the issue of multi-modal integration complexity using the same mapping and similarity calculation for data from different media types. Nonetheless, this may cause information loss during the mapping process due to overlooking the specifics of each individual modality. In this paper, we propose a simple yet effective cross-modal hashing approach, termed collective reconstructive embeddings (CRE), which can simultaneously solve the heterogeneity and integration complexity of multi-modal data. To address the heterogeneity challenge, we propose to process heterogeneous types of data using different modality-specific models. Specifically, we model textual data with cosine similarity-based reconstructive embedding to alleviate the data sparsity to the greatest extent, while for image data, we utilize the Euclidean distance to characterize the relationships of the projected hash codes. Meanwhile, we unify the projections of text and image to the Hamming space into a common reconstructive embedding through rigid mathematical reformulation, which not only reduces the optimization complexity significantly but also facilitates the inter-modal similarity preservation among different modalities. We further incorporate the code balance and uncorrelation criteria into the problem and devise an efficient iterative algorithm for optimization. Comprehensive experiments on four widely used multimodal benchmarks show that the proposed CRE can achieve a superior performance compared with the state of the art on several challenging cross-modal tasks.

Hashing based approximate nearest neighbor search embeds high dimensional data to compact binary codes, which enables efficient similarity search and storage. However, the non-isometry sign() function makes it hard to project the nearest neighbors in continuous data space into the closest codewords in discrete Hamming space. In this work, we revisit the sign() function from the perspective of space partitioning. In specific, we bridge the gap between k-nearest neighbors and binary hashing codes with Shannon entropy. We further propose a novel K-Nearest Neighbors Hashing (KNNH) method to learn binary representations from KNN within the subspaces generated by sign(). Theoretical and experimental results show that the KNN relation is of central importance to neighbor preserving embeddings, and the proposed method outperforms the state-of-the-arts on benchmark datasets.

Cross-media hashing, which maps data from different modalities to a low-dimensional sharing Hamming space, has attracted considerable attention due to the rapid increase of multimodal data, for example, images and texts. Recent cross-media hashing works mainly aim at learning compact hash codes to preserve the class label-based or feature-based similarities among samples. However, these methods ignore the unbalanced semantic gaps between different modalities and high-level semantic concepts, which generally results in less effective hash functions and unsatisfying retrieval performance. Specifically, the key words of texts contain semantic meanings, while the low-level features of images lack of semantic meanings. That means the semantic gap in image modality is larger than that in text modality. In this paper, we propose a simple yet effective hashing method for cross-media retrieval to address this problem, dubbed discrete semantic alignment hashing (DSAH). First, DSAH formulates to exploit collaborative filtering to mine the relations between class labels and hash codes, which can reduce memory consumption and computational cost compared to pairwise similarity. Then, the attribute of image modality is employed to align the semantic information with text modality. Finally, to further improve the quality of hash codes, we propose a discrete optimization algorithm to learn discrete hash codes directly, and each bit has a closed-form solution. Extensive experiments on multiple public databases show that our model can seamlessly incorporate attributes and achieve promising performance.

In recent years, binary code learning, a.k.a hashing, has received extensive attention in large-scale multimedia retrieval. It aims to encode high-dimensional data points to binary codes, hence the original high-dimensional metric space can be efficiently approximated via Hamming space. However, most existing hashing methods adopted offline batch learning, which is not suitable to handle incremental datasets with streaming data or new instances. In contrast, the robustness of the existing online hashing remains as an open problem, while the embedding of supervised/semantic information hardly boosts the performance of the online hashing, mainly due to the defect of unknown category numbers in supervised learning. In this paper, we proposed an online hashing scheme, termed Hadamard Codebook based Online Hashing (HCOH), which aims to solve the above problems towards robust and supervised online hashing. In particular, we first assign an appropriate high-dimensional binary codes to each class label, which is generated randomly by Hadamard codes to each class label, which is generated randomly by Hadamard codes. Subsequently, LSH is adopted to reduce the length of such Hadamard codes in accordance with the hash bits, which can adapt the predefined binary codes online, and theoretically guarantee the semantic similarity. Finally, we consider the setting of stochastic data acquisition, which facilitates our method to efficiently learn the corresponding hashing functions via stochastic gradient descend (SGD) online. Notably, the proposed HCOH can be embedded with supervised labels and it not limited to a predefined category number. Extensive experiments on three widely-used benchmarks demonstrate the merits of the proposed scheme over the state-of-the-art methods. The code is available at https://github.com/lmbxmu/mycode/tree/master/2018ACMMM_HCOH.