图像语义标注中的叙词查询方法*

为了在图像语义标注领域能更好地反映标注之间的关系,通过对已标注图像的标注进行分析来建立标 注之间的关系,并在此基础上将叙词查询的概念引入到图像语义标注中并提出了基于叙词查询的图像语义标注 方法,把语义标注问题统一在叙词查询与图像的语义关系相结合在统一的框架下,最后通过在Ｃｏｒｅｌ图像数据库中的验证表明,所提出的方法是有效的并且标注率得到了明显的提高。     

Query Processing Techniques for Multimedia Presentations

A multimedia presentation is a synchronized, and possibly interactive, delivery of multimedia data to users. We expect that, in the future, multimedia presentations will be stored into and queried from multimedia databases. In an earlier work, we have designed a graphical query language, called GVISUAL, that allows users to query multimedia presentations based on content information. In this paper, we discuss GVISUAL query processing techniques for multimedia presentations. More specifically, we discuss the translation of GVISUAL queries into an operator-based language, called O-Algebra, with three new operators, and efficient implementations of the new O-Algebra operators using a coding system called nodecodes.     

A semantic approach to optimize linear datalog programs

In this work, we present a semantic query optimization approach to improve the efficiency of the evaluation of a subset of SQL:1999 recursive queries. Using datalog notation, we can state our main contribution as an algorithm that builds a program P′ equivalent to a given program P, when both are applied over a database d satisfying a set of functional dependencies. The input program P is a linear recursive datalog program. The new program P′ has less different variables and, sometimes, less atoms in rules, thus it is cheaper to evaluate. Using coral and ibm db2, P′ is empirically shown to be more efficient than the original program.This work is partially supported by Xunta de Galicia grant PGIDIT05SIN10502PR and Ministerio de Educación y Ciencia (PGE y FEDER) grants TIC2003-06593 and TIN2006-15071-C03-03.     

A probabilistic semantic model for image annotation and multi-modal image retrieval

This paper addresses automatic image annotation problem and its application to multi-modal image retrieval. The contribution of our work is three-fold. (1) We propose a probabilistic semantic model in which the visual features and the textual words are connected via a hidden layer which constitutes the semantic concepts to be discovered to explicitly exploit the synergy among the modalities. (2) The association of visual features and textual words is determined in a Bayesian framework such that the confidence of the association can be provided. (3) Extensive evaluation on a large-scale, visually and semantically diverse image collection crawled from Web is reported to evaluate the prototype system based on the model. In the proposed probabilistic model, a hidden concept layer which connects the visual feature and the word layer is discovered by fitting a generative model to the training image and annotation words through an Expectation-Maximization (EM) based iterative learning procedure. The evaluation of the prototype system on 17,000 images and 7736 automatically extracted annotation words from crawled Web pages for multi-modal image retrieval has indicated that the proposed semantic model and the developed Bayesian framework are superior to a state-of-the-art peer system in the literature.     

Symbolic Intersect Detection: A Method for Improving Spatial Intersect Joins

Due to the increasing popularity of spatial databases, researchers have focused their efforts on improving the query processing performance of the most expensive spatial database operation: the spatial join. While most previous work focused on optimizing the filter step, it has been discovered recently that, for typical GIS data sets, the refinement step of spatial join processing actually requires a longer processing time than the filter step. Furthermore, two-thirds of the time in processing the refinement step is devoted to the computation of polygon intersections. To address this issue, we therefore introduce a novel approach to spatial join optimization that drastically reduces the time of the refinement step. We propose a new approach called Symbolic Intersect Detection (SID) for early detection of true hits. Our SID optimization eliminates most of the expensive polygon intersect computations required by a spatial join by exploiting the symbolic topological relationships between the two candidate polygons and their overlapping minimum bounding rectangle. One important feature of our SID optimization is that it is complementary to the state-of-the-art methods in spatial join processing and therefore can be utilized by these techniques to further optimize their performance. In this paper, we also develop an analytical cost model that characterizes SIDs effectiveness under various conditions. Based on real map data, we furthermore conduct an experimental evaluation comparing the performance of the spatial joins with SID against the state-of-the-art approach. Our experimental results show that SID can effectively identify more than 80% of the true hits with negligible overhead. Consequently, with SID, the time needed for resolving polygon intersect in the refinement step is improved by over 50% over known techniques, as predicted by our analytical model.     

基于模糊熵的空间语义图像检索模型研究*

根据模糊熵理论和改进的空间信息分布,提出了颜色空间特征语义图像检索模型。阐述基于语法规则的颜色空间特征语义描述方法,构造从低层颜色空间特征到高层语义之间的映射,根据这些模糊语义值进行图像检索。实验结果表明,该模型能够有效地对图像高层语义进行刻画,由此实现的模型不仅能获得高效和稳定的检索结果,获得与人类视觉感知较好的一致性,该算法还能很好地消除低层图像空间特征和高层语义之间的语义鸿沟。     

部件检测和语义网络的细粒度鞋类图像检索

目的 细粒度图像检索是当前细粒度图像分析和视觉领域的热点问题。以鞋类图像为例,传统方法仅提取其粗粒度特征且缺少关键的语义属性,难以区分部件间的细微差异,不能有效用于细粒度检索。针对鞋类图像检索大多基于简单款式导致检索效率不高的问题,提出一种结合部件检测和语义网络的细粒度鞋类图像检索方法。方法 结合标注后的鞋类图像训练集对输入的待检鞋类图像进行部件检测;基于部件检测后的鞋类图像和定义的语义属性训练语义网络,以提取待检图像和训练图像的特征向量,并采用主成分分析进行降维;通过对鞋类图像训练集中每个候选图像与待检图像间的特征向量进行度量学习,按其匹配度高低顺序输出检索结果。结果 实验在UT-Zap50K数据集上与目前检索效果较好的4种方法进行比较,检索精度提高近6%。同时,与同任务的SHOE-CNN（semantic hierarchy of attribute convolutional neural network）检索方法比较,本文具有更高的检索准确率。结论 针对传统图像特征缺少细微的视觉描述导致鞋类图像检索准确率低的问题,提出一种细粒度鞋类图像检索方法,既提高了鞋类图像检索的精度和准确率,又能较好地满足实际应用需求。     

A multiscale approach to texture-based image retrieval

This paper presents research on a robust technique for texture-based image retrieval in multimedia museum collections. The aim is to be able to use a query image patch containing a single texture to retrieve images containing an area with similar texture to that in the query. The feature extractor used to build the feature vectors is based on an improved version of the discrete wavelet frames (DWF), proposed elsewhere. In order to utilise the feature extractor on real scene image datasets, a block-oriented decomposition technique, termed the multiscale sub-image matching method, is presented. The multiscale method, together with the DWF, provide an efficient content-based retrieval technique without the need for segmentation. The algorithms are tested on a range of databases of texture images as well as on real museum image collections. Promising results are reported.

A framework for semantic grouping in P2P databases

Sharing of structured data in decentralized environments is a challenging problem, especially in the absence of a global schema. Social network structures map network links to semantic relations between participants in order to assist in efficient resource discovery and information exchange. In this work, we propose a scheme that automates the process of creating schema synopses from semantic clusters of peers which own autonomous relational databases. The resulting mediated schemas can be used as global interfaces for relevant queries. Active nodes are able to initiate the group schema creation process, which produces a mediated schema representative of nodes with similar semantics. Group schemas are then propagated in the overlay and used as a single interface for relevant queries. This increases both the quality and the quantity of the retrieved answers and allows for fast discovery of interest groups by joining peers. As our experimental evaluations show, this method increases both the quality and the quantity of the retrieved answers and allows for faster discovery of semantic groups by joining peers.     

面向Web图像检索的基于语义迁移的无监督深度哈希

当前主流的Web图像检索方法仅考虑了视觉特征,没有充分利用Web图像附带的文本信息,并忽略了相关文本中涉及的有价值的语义,从而导致其图像表达能力不强。针对这一问题,提出了一种新的无监督图像哈希方法——基于语义迁移的深度图像哈希（semantic transfer deep visual hashing,STDVH）。该方法首先利用谱聚类挖掘训练文本的语义信息;然后构建深度卷积神经网络将文本语义信息迁移到图像哈希码的学习中;最后在统一框架中训练得到图像的哈希码和哈希函数,在低维汉明空间中完成对大规模Web图像数据的有效检索。通过在Wiki和MIR Flickr这两个公开的Web图像集上进行实验,证明了该方法相比其他先进的哈希算法的优越性。     

基于地理空间本体的语义检索相关度研究

以地理信息领域为应用背景,面向地理空间语义检索,基于地球信息科学中的空间拓扑理论,以空间本体为语义检索的概念空间,提出了一种语义相关度的算法。其特点是考虑了传统字面匹配相关度与语义关系相关度两部分的融合,同时引入了本体关系权值的机制控制在不同语义检索应用中本体的关联程度,并体现了其与语义距离的反比关系。通过所作的相关实验,验证了该语义相关度算法在地理空间语义检索应用中可以达到良好的效果,并且也为其他领域应用提供了较好的参考和借鉴价值。     

A scalable approach for content based image retrieval in cloud datacenter

The emergence of cloud datacenters enhances the capability of online data storage. Since massive data is stored in datacenters, it is necessary to effectively locate and access interest data in such a distributed system. However, traditional search techniques only allow users to search images over exact-match keywords through a centralized index. These techniques cannot satisfy the requirements of content based image retrieval (CBIR). In this paper, we propose a scalable image retrieval framework which can efficiently support content similarity search and semantic search in the distributed environment. Its key idea is to integrate image feature vectors into distributed hash tables (DHTs) by exploiting the property of locality sensitive hashing (LSH). Thus, images with similar content are most likely gathered into the same node without the knowledge of any global information. For searching semantically close images, the relevance feedback is adopted in our system to overcome the gap between low-level features and high-level features. We show that our approach yields high recall rate with good load balance and only requires a few number of hops.     

A new matching strategy for content based image retrieval system

Adopting effective model to access the desired images is essential nowadays with the presence of a huge amount of digital images. The present paper introduces an accurate and rapid model for content based image retrieval process depending on a new matching strategy. The proposed model is composed of four major phases namely: features extraction, dimensionality reduction, ANN classifier and matching strategy. As for the feature extraction phase, it extracts a color and texture features, respectively, called color co-occurrence matrix (CCM) and difference between pixels of scan pattern (DBPSP). However, integrating multiple features can overcome the problems of single feature, but the system works slowly mainly because of the high dimensionality of the feature space. Therefore, the dimensionality reduction technique selects the effective features that jointly have the largest dependency on the target class and minimal redundancy among themselves. Consequently, these features reduce the calculation work and the computation time in the retrieval process. The artificial neural network (ANN) in our proposed model serves as a classifier so that the selected features of query image are the input and its output is one of the multi classes that have the largest similarity to the query image. In addition, the proposed model presents an effective feature matching strategy that depends on the idea of the minimum area between two vectors to compute the similarity value between a query image and the images in the determined class. Finally, the results presented in this paper demonstrate that the proposed model provides accurate retrieval results and achieve improvement in performance with significantly less computation time compared with other models.     

Using extended feature objects for partial similarity retrieval

In this paper, we introduce the concept of extended feature objects for similarity retrieval. Conventional approaches for similarity search in databases map each object in the database to a point in some high-dimensional feature space and define similarity as some distance measure in this space. For many similarity search problems, this feature-based approach is not sufficient. When retrieving partially similar polygons, for example, the search cannot be restricted to edge sequences, since similar polygon sections may start and end anywhere on the edges of the polygons. In general, inherently continuous problems such as the partial similarity search cannot be solved by using point objects in feature space. In our solution, we therefore introduce extended feature objects consisting of an infinite set of feature points. For an efficient storage and retrieval of the extended feature objects, we determine the minimal bounding boxes of the feature objects in multidimensional space and store these boxes using a spatial access structure. In our concrete polygon problem, sets of polygon sections are mapped to 2D feature objects in high-dimensional space which are then approximated by minimal bounding boxes and stored in an R-tree. The selectivity of the index is improved by using an adaptive decomposition of very large feature objects and a dynamic joining of small feature objects. For the polygon problem, translation, rotation, and scaling invariance is achieved by using the Fourier-transformed curvature of the normalized polygon sections. In contrast to vertex-based algorithms, our algorithm guarantees that no false dismissals may occur and additionally provides fast search times for realistic database sizes. We evaluate our method using real polygon data of a supplier for the car manufacturing industry. Edited by R. Güting. Received October 7, 1996 / Accepted March 28, 1997     

Modeling continuous visual features for semantic image annotation and retrieval

Automatic image annotation has become an important and challenging problem due to the existence of semantic gap. In this paper, we firstly extend probabilistic latent semantic analysis (PLSA) to model continuous quantity. In addition, corresponding Expectation-Maximization (EM) algorithm is derived to determine the model parameters. Furthermore, in order to deal with the data of different modalities in terms of their characteristics, we present a semantic annotation model which employs continuous PLSA and standard PLSA to model visual features and textual words respectively. The model learns the correlation between these two modalities by an asymmetric learning approach and then it can predict semantic annotation precisely for unseen images. Finally, we compare our approach with several state-of-the-art approaches on the Corel5k and Corel30k datasets. The experiment results show that our approach performs more effectively and accurately.     

Color image retrieval using multispectral random field texture model and color content features

This paper describes a color-texture-based image retrieval system for query of an image database to find similar images to a target image. The color-texture information is obtained via modeling with the multispectral simultaneous autoregressive (MSAR) random field model. The general color content characterized by ratios of sample color means is also used. The retrieval process involves segmenting the image into regions of uniform color texture using an unsupervised histogram clustering approach that utilizes the combination of MSAR and color features. The color-texture content, location, area and shape of the segmented regions are used to develop similarity measures describing the closeness of a query image to database images. These attributes are derived from the maximum fitting square and best fitting ellipse to each of the segmented regions. The proposed similarity measure combines all these attributes to rank the closeness of the images. The performance of the system is tested on two databases containing synthetic mosaics of natural textures and natural scenes, respectively.