High-speed Retrieval Research Articles

Research on the Construction of Statistical Data Management Platform under Big Data

Purpose: Under the construction of Statistical data management platform, the problems such as low accuracy of data, difficult effective integration of data, low efficiency, etc., are solved, and the clinical research workflow is improved. Methods: Based on the integration platform and big data related technologies, multi-source heterogeneous data are integrated and cleaned to form a scientific research data center. What's more, the scientific research retrieval system is used for high-speed retrieval, analysis and utilization of the data in the scientific research data center. Results: The built scientific research retrieval system has the functions of integrating statistical methods, flexibly configuring statistical indicators, customizing queries, etc., which reduces the pressure of scientific researchers in data collection, statistics and analysis, and improves the statistical analysis quality of scientific research data. Conclusion: It has improved the quality and efficiency of scientific research, accelerated the transformation of scientific research achievements, and indirectly promoted the improvement of medical quality.

Deep global semantic structure-preserving hashing via corrective triplet loss for remote sensing image retrieval

With the explosive increase of remote sensing data, how to search for remote sensing data quickly and accurately in a vast dataset is an incredibly critical matter for research subjects. The deep hashing method has become the dominant method for remote sensing image retrieval because of its low-cost storage and high-speed retrieval. However, for the reason of the limitation of fixed convolutional kernels, deep hashing frameworks based on Convolutional Neural Networks (CNNs) fail to obtain the global semantic information well, which leads to the generation of suboptimal solutions. Furthermore, existing hashing methods commonly employ the random sampling strategy or hardest sample mining to build training batches, resulting in bad local minima. To remedy these problems, a novel Deep Global Semantic Structure-preserving Hashing framework via corrective triplet loss (DGSSH) is proposed for remote sensing image retrieval to learn a discriminative and stable embedding space, achieving intra-class confusion and inter-class diversity. Specifically speaking, the feature extraction module based on Swim Transformer architecture is developed to capture global semantic information and multiscale features from remote sensing images. Based on a distribution matching constraint, the corrective triplet loss for deep hashing schemes is designed to reduce the distribution shift caused by the random selection or hardest sample mining. Meanwhile, to reduce the time overhead of the model, the asymmetric learning strategy is employed to perform effective compact representation learning. Numerous experiments have been carried out on three publicly available benchmarks, which indicates that the proposed DGSSH framework could achieve optimal performance for remote sensing image retrieval applications. The source code of our DGSSH framework is hosted at https://github.com/QinLab-WFU/DGSSH.git.

Landscape Perspective Distance-Included Angle Shape Distribution Analysis Based on 3D CAD Model Retrieval Algorithm

In view of the results obtained in the retrieval process of the 3D CAD model, which can show the differences in the local feature details of the model, the 3D CAD model retrieval algorithm is introduced into the analysis of the perspective distance-angle shape distribution of the garden landscape in this paper. Random sampling is performed on the surface of the constructed 3D CAD model, combined with the test distance between the sampling point and the neighboring points, and the corresponding garden landscape perspective distance-angle shape distribution characteristics in this area are calculated in order to achieve the similarity of the CAD model high-speed retrieval. Finally, experimental research shows that the algorithm proposed in this paper is better than the overall shape distribution algorithm and the spherical harmonic algorithm in the search performance of the CAD model, and it can effectively improve the recognition ability of the local detailed features of the 3D CAD model.

Massive picture retrieval system based on big data image mining

The traditional picture retrieval system has a slow retrieval speed, poor retrieval accuracy, and a low recall when performing massive picture retrieval. In this paper, we design a massive picture retrieval system using the big data image mining technology. It is constructed with data processing layer, business logic layer and presentation layer and works through three steps of data segmentation, mining and merging. For instance, it runs the distributed file system module in a Master/Slave operation mode and designs file read and write requests according to user interaction. Next, it performs parallel computing of picture data sets based on Map Reduce module to solve the picture matching and similarity metrics and returns to the user sorted picture matching result. Then, it extracts the color and texture features of the target area to generate the final picture retrieval result. We select a large number of pictures on a big data platform as simulation test set. The results show that the system we designed has a good retrieval accuracy and a high retrieval speed, which greatly improves the recall of picture retrieval.

Deep Multi-View Enhancement Hashing for Image Retrieval.

Hashing is an efficient method for nearest neighbor search in large-scale data space by embedding high-dimensional feature descriptors into a similarity preserving Hamming space with a low dimension. However, large-scale high-speed retrieval through binary code has a certain degree of reduction in retrieval accuracy compared to traditional retrieval methods. We have noticed that multi-view methods can well preserve the diverse characteristics of data. Therefore, we try to introduce the multi-view deep neural network into the hash learning field, and design an efficient and innovative retrieval model, which has achieved a significant improvement in retrieval performance. In this paper, we propose a supervised multi-view hash model which can enhance the multi-view information through neural networks. This is a completely new hash learning method that combines multi-view and deep learning methods. The proposed method utilizes an effective view stability evaluation method to actively explore the relationship among views, which will affect the optimization direction of the entire network. We have also designed a variety of multi-data fusion methods in the Hamming space to preserve the advantages of both convolution and multi-view. In order to avoid excessive computing resources on the enhancement procedure during retrieval, we set up a separate structure called memory network which participates in training together. The proposed method is systematically evaluated on the CIFAR-10, NUS-WIDE and MS-COCO datasets, and the results show that our method significantly outperforms the state-of-the-art single-view and multi-view hashing methods.

Deep Semantic-Preserving Reconstruction Hashing for Unsupervised Cross-Modal Retrieval.

Deep hashing is the mainstream algorithm for large-scale cross-modal retrieval due to its high retrieval speed and low storage capacity, but the problem of reconstruction of modal semantic information is still very challenging. In order to further solve the problem of unsupervised cross-modal retrieval semantic reconstruction, we propose a novel deep semantic-preserving reconstruction hashing (DSPRH). The algorithm combines spatial and channel semantic information, and mines modal semantic information based on adaptive self-encoding and joint semantic reconstruction loss. The main contributions are as follows: (1) We introduce a new spatial pooling network module based on tensor regular-polymorphic decomposition theory to generate rank-1 tensor to capture high-order context semantics, which can assist the backbone network to capture important contextual modal semantic information. (2) Based on optimization perspective, we use global covariance pooling to capture channel semantic information and accelerate network convergence. In feature reconstruction layer, we use two bottlenecks auto-encoding to achieve visual-text modal interaction. (3) In metric learning, we design a new loss function to optimize model parameters, which can preserve the correlation between image modalities and text modalities. The DSPRH algorithm is tested on MIRFlickr-25K and NUS-WIDE. The experimental results show that DSPRH has achieved better performance on retrieval tasks.

Flexible Cross-Modal Hashing.

Hashing has been widely adopted for large-scale data retrieval in many domains due to its low storage cost and high retrieval speed. Existing cross-modal hashing methods optimistically assume that the correspondence between training samples across modalities is readily available. This assumption is unrealistic in practical applications. In addition, existing methods generally require the same number of samples across different modalities, which restricts their flexibility. We propose a flexible cross-modal hashing approach (FlexCMH) to learn effective hashing codes from weakly paired data, whose correspondence across modalities is partially (or even totally) unknown. FlexCMH first introduces a clustering-based matching strategy to explore the structure of each cluster and, thus, to find the potential correspondence between clusters (and samples therein) across modalities. To reduce the impact of an incomplete correspondence, it jointly optimizes the potential correspondence, the cross-modal hashing functions derived from the correspondence, and a hashing quantitative loss in a unified objective function. An alternative optimization technique is also proposed to coordinate the correspondence and hash functions and reinforce the reciprocal effects of the two objectives. Experiments on public multimodal data sets show that FlexCMH achieves significantly better results than state-of-the-art methods, and it, indeed, offers a high degree of flexibility for practical cross-modal hashing tasks.

Cloud Storage Data Retrieval Method of Port Logistic Based on Particle Swarm Optimization

Qin, Y.Y., 2020. Cloud storage data retrieval method of port logistic based on particle swarm optimization. In: Al-Tarawneh, O. and Megahed, A. (eds.), Recent Developments of Port, Marine, and Ocean Engineering. Journal of Coastal Research, Special Issue No. 110, pp. 295–298. Coconut Creek (Florida), ISSN 0749-0208.The arrival of big data is a major challenge for information technology. In particular, cloud storage and retrieval of big data is also an important technology. It is also an important part of national information security. The cloud storage data retrieval method of port logistic based on particle swarm optimization algorithm proposed in this paper, combined with Matlab software, realizes the PSO algorithm optimization model, and establishes an experimental scenario to simulate the retrieval process of cloud storage. The results prove that the data retrieval method of port logistic has a high retrieval speed and stability. In the future research, the author will focus on the experimental improvement of the method of port logistic to obtain better application effect in the data retrieval process of cloud storage.

Efficient Parameter-Free Adaptive Multi-Modal Hashing

Unsupervised multi-modal hashing has recently attracted broad attention in research area of large-scale multimedia retrieval for its low storage cost, high retrieval speed, and independence on semantic labels. However, the model learning process of existing methods still suffer from the problem of low efficiency: 1) Many existing methods measure the contributions of different modalities using fixed modality weights. In order to avoid over-fitting, they need an inefficient hyper-parameter adjustment process. 2) Most existing methods adopt inefficient optimization strategies to learn hash codes. In this letter, we propose an unsupervised Efficient Parameter-free Adaptive Multi-modal Hashing (EPAMH) model to adaptively capture the modality variations and preserve the discriminative semantics of multi-modal features into the binary hash codes. Moreover, we directly learn the binary codes with simple and efficient operations, which prevents the relaxing quantization errors and improves the model learning efficiency. Experiments prove the superior performance of EPAMH on three public multimedia retrieval datasets. Our source codes and testing datasets can be obtained at https://github.com/ChaoqunZheng/EPAMH .

Index-Based Searching of Interaction Patterns in Large Collections of Protein-Ligand Interfaces.

Comparison of three-dimensional interaction patterns in large collections of protein-ligand interfaces is a key element for understanding protein-ligand interactions and supports various steps in the structure-based drug design process. Different methods exist that provide query systems to search for geometrical patterns in protein-ligand complexes. However, these tools do not meet all of the requirements, which are high query variability, an adjustable search set, and high retrieval speed. Here we present a new tool named PELIKAN that is able to search for a variety of geometrical queries in large protein structure collections in a reasonably short time. The data are stored in an SQLite database that can easily be constructed from any set of protein-ligand complexes. We present different test queries demonstrating the performance of the PELIKAN approach. Furthermore, two application scenarios show the usefulness of PELIKAN in structure-based design endeavors.

Automatic Extraction of Objects and their Attributes from Semi-Structured Web Tables for E-commerce Tasks

Most business Web documents provide key information for decision-making in the form of text tables. High-speed retrieval and analysis of such tables is of great interest to the business and at the same time is a major challenge for researchers. The problem lies in the fact that text tables in Web documents are rarely self-described, i.e. contain no data schemas. In this paper we consider the special case of the problem, limited by the scope of e-commerce. The main objects of e-commerce are goods, works, services and other objects represented in text tables as sets of characteristics (attributes). In fact, these text tables compactly map domain objects into sets of “attribute-value” pairs. Thus, the extraction of meaningful information from the tables can be interpreted as extraction of sets of “attribute-value” pairs. We need to interpret these attributes correctly using a domain knowledge base, since the search engine does not have any information about neither the nature of these attributes, nor their internal relations. In this paper we use as knowledge base semi-extensible e-commerce ontology, based on economic classifiers of the Kazakhstan unified system of classification and coding of technical, economic and social information.

A common short-term memory retrieval rate may describe many cognitive procedures

We examine the relationship between response speed and the number of items in short-term memory (STM) in four different paradigms and find evidence for a similar high-speed processing rate of about 25–30 items per second (∼35–40 ms/item). We propose that the similarity of the processing rates across paradigms reflects the operation of a very basic covert memory process, high-speed retrieval, that is involved in both the search for information in STM and the reactivation or refreshing of information that keeps it in STM. We link this process to a specific pattern of rhythmic, repetitive neural activity in the brain (gamma oscillations). This proposal generates ideas for research and calls for an integrative approach that combines neuroscientific measures with behavioral cognitive techniques.

A robust/fast spoken term detection method based on a syllable n-gram index with a distance metric

For spoken document retrieval, it is crucial to consider Out-of-vocabulary (OOV) and the mis-recognition of spoken words. Consequently, sub-word unit based recognition and retrieval methods have been proposed. This paper describes a Japanese spoken term detection method for spoken documents that robustly considers OOV words and mis-recognition. To solve the problem of OOV keywords, we use individual syllables as the sub-word unit in continuous speech recognition. To address OOV words, recognition errors, and high-speed retrieval, we propose a distant n-gram indexing/retrieval method that incorporates a distance metric in a syllable lattice. When applied to syllable sequences, our proposed method outperformed a conventional DTW method between syllable sequences and was about 100 times faster. The retrieval results show that we can detect OOV words in a database containing 44h of audio in less than 10msec per query with an F-measure of 0.54.

Hierarchical Approximate Matching for Retrieval of Chinese Historical Calligraphy Character

As historical Chinese calligraphy works are being digitized, the problem of retrieval becomes a new challenge. But, currently no OCR technique can convert calligraphy character images into text, nor can the existing Handwriting Character Recognition approach does not work for it. This paper proposes a novel approach to efficiently retrieving Chinese calligraphy characters on the basis of similarity: calligraphy character image is represented by a collection of discriminative features, and high retrieval speed with reasonable effectiveness is achieved. First, calligraphy characters that have no possibility similar to the query are filtered out step by step by comparing the character complexity, stroke density and stroke protrusion. Then, similar calligraphy characters are retrieved and ranked according to their matching cost produced by approximate shape match. In order to speed up the retrieval, we employed high dimensional data structure -- PK-tree. Finally, the efficiency of the algorithm is demonstrated by a preliminary experiment with 3012 calligraphy character images.

A retrieval method of a double array in secondary storage

AbstractThe trie method is widely used as a dictionary retrieval technique in natural language processing, and the double array method is extremely good in terms of high retrieval speed and compactness as a data structure for implementing the trie method. However, although placing the double array in main storage is effective for fast retrieval, the size of the double array increases as the number of storage keys increases. Therefore, in this paper, the authors propose a divided double array method that can execute retrieval efficiently even when using a small main storage area and can also further reduce the amount of storage by dividing the double array and storing it in secondary storage. However, a problem with this technique is that the number of accesses to secondary storage increases because the divided tries must be scanned. To deal with this problem, the authors proposed a technique that reduces the number of times the divided tries are scanned to maintain the fast access speed of the double array. They performed experiments for sets containing from 70,000 to 230,000 keys, and from the results, it was apparent that the number of accesses to secondary storage during retrieval for a single key was less than or equal to 2, and the storage capacity required for the divided tries was reduced by approximately 50% compared with the undivided trie. © 2004 Wiley Periodicals, Inc. Syst Comp Jpn, 35(8): 72–81, 2004; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/scj.10481

A filtering method for high-speed retrieval of similar active sites

It is becoming clear that the function of protein is activated by a local part of its molecular surface called an active site. It is impractical to compare the input surface with all of the stored active site data, because thousands of active site data whose functions are known have been stored in the active site DB. This paper proposes a method of active site data filtering, in which only the limited active site data that are expected to be similar to the input protein are roughly filtered for the high-speed retrieval. In this method, characteristic regions, which are defined as parts of the surface that show a similar property, are extracted based on the idea that some similar regions in property and shape are observed in a set of similar protein surfaces. A similarity score between input protein and an active site is roughly calculated by using characteristic regions extracted from each of them. The method was applied to a set of 556 enzyme proteins. The effectiveness of the method was confirmed by the fact that it can shorten the retrieval time to 7.2% of the total retrieval for 90.2% input proteins.

Fast Retrieval of Isolated Visual Shapes

Similarity-based retrieval from databases of isolated visual shapes has become an important information retrieval problem. The goal of the current work is to achieve high retrieval speed with reasonable retrieval effectiveness, and support for partial and occluded shape queries. In the proposed method, histograms of local shape parts are coded as index vectors. To increase retrieval accuracy, a rich set of parts at all scales of the shape is used; specifically, the parts are defined as connected sequences of regions in curvature scale space. To increase efficiency, structural indexing is used to compare the index vectors of the query and database shapes. In experimental evaluations, the method retrieved at least one similar shape in the top three retrieved items 99–100% of the time, depending on the database. Average retrieval times ranged from 0.7 ms on a 131-shape database to 7 ms on a 1310-shape database. The method is thus suitable for fast, approximate shape retrieval in comparison with more accurate but more costly structural matching.

Information retrieval method from uncertain requests and its optical implementation

We propose a high-speed information retrieval method which is tolerant to a user's uncertain request. To consider the degree of certainty of user's remembrance, a user's query is represented as a set of keywords and certainty factor pairs. Optical matrix-vector multiplication is applied to the computation of the retrieval status values to realize a high-speed retrieval system. An experimental result is also presented in this paper.

8.医用画像処理および記憶システムの国内外の開発状況について(I.放射線施設におけるコンピュータの利用について)

Today, hospitals are storing and utilizing medical image data manually. Moreover, because of new imaging diagnostic equipments, the use of image data has greatly increased resulting in too much tasks of storing, transporting and retrieving this data, therefore causing many troublesome problems for the function and management of hospitals. To solve these problems, TOSHIBA has developed TDIS, Total Digital Imaging System, which realizes the effective and economical performance of imaging diagnosis by incorporating many latest techniques ; sensing, collecting, mass storage, high speed retrieval, fast transmission, data process and high resolution display of medical images. TDIS is contained in the field of the technique on Picture Archiving and Communication System so called PACS. Annually International Conference of Picture Archiving and Communicaton for Medical Applications is held, and many engineers and doctors from many kinds of fields participate in research and development of these systems. A stationary card optical data storage device is described. Digital x-ray images were stored and retrieved at high data rates. Average density can be approximately 3×10^3 bits per in^2. A card has certain advantages for medical data archiving.