Knowing the location where a photograph was taken provides us with data that could be useful in a wide spectrum of applications. With the advance of digital cameras, and with many users exchanging their digital cameras for GPS-enabled mobile phones, photographs annotated with geographical locations are becoming ever more present on photo-sharing websites such as Flickr. However there is still a mass of content that is not geotagged, meaning that algorithms for efficient and accurate geographical estimation of an image are needed. This paper presents a general model for effectively using both textual metadata and visual features of photos to automatically place them on a world map with state-of-the-art performance. In addition, we explore how information from user-modelling can be fused with our model, and investigate the effect such modelling has on performance.

Search result diversification can increase user satisfaction in answering a particular information need. There are many ways of diversify search results. In some cases the user has a clear idea of how they would like to see their results diversified. This work presents a system that is capable of diversifying search results along specific user-specified axes of diversity.

Millions of images are tweeted every day, yet very little research has looked at the non-textual aspect of social media communication. In this work we have developed a system to analyse streams of image data. In particular we explore trends in similar, related, evolving or even duplicated visual artefacts in the mass of tweeted image data - in short, we explore the visual pulse of Twitter.

ImageTerrier is a novel easily extensible open-source, scalable, high-performance search engine platform for content-based image retrieval applications. The platform provides a comprehensive test-bed for experimenting with bag-of-visual-words image retrieval techniques. It incorporates a state-of-the-art implementation of the single-pass indexing technique for constructing inverted indexes and is capable of producing highly compressed index data structures. ImageTerrier is written as an extension to the open-source Terrier, "Terabyte Retriever", test-bed platform for textual information retrieval research. The ImageTerrier platform is demonstrated to successfully index and search a corpus of over 10 million images containing just under 10,000,000,000 quantised SIFT visual terms.

The SIFT keypoint descriptor is a powerful approach to encoding local image description using edge orientation histograms. Through codebook construction via k-means clustering and quantisation of SIFT features we can achieve image retrieval treating images as bags-of-words. Intensity inversion of images results in distinct SIFT features for a single local image patch across the two images. Intensity inversions notwithstanding these two patches are structurally identical. Through careful reordering of the SIFT feature vectors, we can construct the SIFT feature that would have been generated from a non-inverted image patch starting with those extracted from an inverted image patch. Furthermore, through examination of the local feature detection stage, we can estimate whether a given SIFT feature belongs in the space of inverted features, or non-inverted features. Therefore we can consistently separate the space of SIFT features into two distinct subspaces. With this knowledge, we can demonstrate reduced time complexity of codebook construction via clustering by up to a factor of four and also reduce the memory consumption of the clustering algorithms while producing equivalent retrieval results.

With the increasing amount of multimedia content on the web added as user generated content in Web 2.0 websites, conventional multimedia information retrieval is presented with new challenges. It is no longer possible to rely only on meta-data based retrieval but to consider also content based techniques combined with the collective knowledge generated by users' contributions and geo-referenced meta-data. Tagging is a modest way to annotate such documents and fails to capture a full semantic description of the document content. This report concerns ongoing research to investigate a means to identify, model and utilise semantic descriptions of the user-generated content in Web 2.0 documents using a hybrid approach. The approach consists of three main components, natural language processing, image analysis and a shared knowledge base. In this paper we describe the complete model but, as the image analysis component is in its early stages, the results focus on the natural language processing and the knowledge base. We show that the additional use of these components can improve retrieval and analysis performance over that based only on Web 2.0 tags.

The current trend in distributed systems is towards service-based integration. This article describes an ontology-driven framework implemented to provide knowledge management for data of different modalities, with multimedia processing, annotation, and reasoning provided by remote services. The framework was developed in, and is presented in the context of, the Medical Imaging and Advanced Knowledge Technologies (MIAKT) project that sought to support the Multidisciplinary Meetings (MDMs) that take place during breast cancer screening for diagnosing the patient. However, the architecture is entirely independent of the specific application domain and can be quickly prototyped into new domains. An Enterprise server provides resource access to a client-side presentation application which, in turn, provides knowledge visualization and markup of any supported media, as defined by a domain-dependent ontology-supported language.

A distributed decision support system involving multiple clinical centres is crucial to the diagnosis of rare diseases. Although sharing of valid diagnosed cases can facilitate later decision making, possibly from geographically different centres, the released information could reveal patient privacy if it is not properly protected. Clinical centres may have to impose their distinct regulations and rules that govern the use of their data externally. The collaboration of centres, therefore, must respect the collective policies and ideally, serve users the most appropriate and useful resources possible in the system according to the past experience. In this way, the system’s value is entrusted and even elevated through continuous collaboration. We present in this paper a link-anonymised data scheme and in addition to that, a security model that together enforce privacy data security and secure resource access for distributed clinical centres. Our illustration of the approach involves a prototype medical decision support system, HealthAgents, for brain tumour diagnosis.

In this paper we introduce a new graph based bidding language for combinatorial auctions. In our language, each bidder submits to the arbitrator a generalized flow network (netbid) representing her bids. The interpretation of the winner determination problem as an aggregation of individual preferences represented as flowbids allows building an aggregate netbid for its representation. Labelling the nodes with appropriate procedural functions considerably improves upon the expressivity of previous bidding languages.

In this paper, we describe the unique security issues involved in healthcare domains. These have been addressed to the needs of the HealthAgents project. In the proposed approach, several levels of security have been provided in accordance with Software Engineering principles, ethical regulations for healthcare data, as well as the security requirements usually raised from the distributed clinical settings. The result is the production of a secure and maintainable Multi-Agent System that enables secure communication, uniform home site authentication, and customised resource access authorisation. A security policy rule scheme has been designed for agent interaction modelling. This separates the functional and non-functional (security) requirements but let security policy constraints integrate into the running of the agents via a unified role notion. Each user/agent can play a function role only when its assigned social rights roles permit the access to resources of various types and geographical locations, as specified in the function role behaviour. The approach is illustrated using a comprehensive secure access case.

This paper focuses on the problem of representing, in a meaningful way, the knowledge involved in the HealthAgents project. Our work is motivated by the complexity of representing electronic healthcare records in a consistent manner. We present HADOM (HealthAgents domain ontology) which conceptualises the required HealthAgents information and propose describing the sources knowledge by the means of conceptual graphs (CGs). This allows to build upon the existing ontology permitting for modularity and flexibility. The novelty of our approach lies in the ease with which CGs can be placed above other formalisms and their potential for optimised querying and retrieval.

We present in this paper an adaptive security model for Multi-agent systems. A security meta-model has been developed in which the traditional role concept has been extended. The new concept incorporates the need of both security management as used by role-based access control (RBAC) and agent functional behaviour in agent-oriented Software Engineering (AOSE). Our approach avoids weaknesses of traditional RBAC approaches and provides a practically usable security model for multi-agent systems (MAS). A unified role interaction model framework has been put forward that incorporates not only functional requirements but also security constraints in MAS. A security policy rule scheme has been used to express security requirements in relation to affective roles. The major contribution of the work is that little redevelopment effort will be required when security is to be engineered into the overall MAS architecture, hence minimising the impact of the security requirements changes to the MAS architecture. We illustrate the approach through its potential application in a clinical trial setting involving a prototype medical decision support system, HealthAgents.

Software systems must be capable of coping with continuous requirements changes and at the same time wisely make use of emerging components and services to remain useful in their environment. In this paper, the adaptive agent model (AAM) approach is proposed. The AAM uses configurable interaction model to drive adaptive agent behaviour. The model captures user requirements and is maintained by experts at a high level of abstraction. The AAM interaction model has been discussed with regard to interaction specification and interaction coordination, in line with a coordination language for the OpenKnowledge project. A major benefit of using the approach is agents can dynamically choose disparate components and services already developed for computation via their interaction with each other at runtime, when a new interaction model has been configured for them towards an emerging business goal. A simple expert seeking scenario has been used to illustrate the approach.

Ontologies, as a knowledge representation paradigm, rely on semantics for performing logical inference. However, in the context of ontology query and reuse, the semantics are hard to be manipulated. When searching for ontologies and identifying potentially relevant candidates, a mechanism is required to exploit (i) the appropriate semantic restrictions as well as the (ii) meaning of concepts/properties. This paper proposes an approach that will address both requirements. The contribution of this paper is to conceptually detail how a set of ontology indices can be obtained. More specifically, to obtain a relevant set of keywords describing an ontology, we use logic-based constructors to generate an initial set of keywords. The meaning of these keywords is further explored by consulting a comprehensive text corpora serving as the background/domain knowledge. The novelty of our work lies in the combination of logic-based formalisms with a collaboratively contributed web repository. We illustrate our work by the means of a real world example.

The Corel Image set is widely used for image annotation performance evaluation although it has been claimed that Corel images are relatively easy to annotate. The aim of this paper is to demonstrate some of the disadvantages of datasets like the Corel set for effective auto-annotation evaluation. We first compare the performance of several annotation algorithms using the Corel set and find that simple near neighbor propagation techniques perform fairly well. A support vector machine (SVM)-based annotation method achieves even better results, almost as good as the best found in the literature. We then build a new image collection using the Yahoo Image Search engine and query-by-single-word searches to create a more challenging annotated set automatically. Then, using three very different image annotation methods, we demonstrate some of the problems of annotation using the Corel set compared with the Yahoo-based training set. In both cases the training sets are used to create a set of annotations for the Corel test set

Traditionally, statistical models for image auto-annotation have been coupled with image segmentation. Considering the performance of the current segmentation algorithms, it can be meaningful to avoid a segmentation stage. In this paper, we propose a new approach to image auto-annotation by building on previously developed statistical models. In this approach, segmentation is avoided through the use of salient regions. The use of the statistical model results in an annotation performance which improves upon our previously proposed saliency-based word propagation technique. We also show that the use of salient regions achieves better results than the use of general image regions or segments

In this paper, we describe research into the use of ontologies to integrate access to cultural heritage and photographic archives. The use of the CIDOC CRM and CRM Core ontologies are described together with the metadata mapping methodology. A system integrating data from four content providers will be demonstrated.

A new approach to image retrieval is presented in the domain of museum and gallery image collections. Specialist algorithms, developed to address specific retrieval tasks, are combined with more conventional content and metadata retrieval approaches, and implemented within a distributed architecture to provide cross-collection searching and navigation in a seamless way. External systems can access the different collections using interoperability protocols and open standards, which were extended to accommodate content based as well as text based retrieval paradigms. After a brief overview of the complete system, we describe the novel design and evaluation of some of the specialist image analysis algorithms including a method for image retrieval based on sub-image queries, retrievals based on very low quality images and retrieval using canvas crack patterns. We show how effective retrieval results can be achieved by real end-users consisting of major museums and galleries, accessing the distributed but integrated digital collections.

This paper presents a fully unsupervised texture segmentation algorithm by using a modified discrete wavelet frames decomposition and a mean shift algorithm. By fully unsupervised, we mean the algorithm does not require any knowledge of the type of texture present nor the number of textures in the image to be segmented. The basic idea of the proposed method is to use the modified discrete wavelet frames to extract useful information from the image. Then, starting from the lowest level, the mean shift algorithm is used together with the fuzzy c-means clustering to divide the data into an appropriate number of clusters. The data clustering process is then refined at every level by taking into account the data at that particular level. The final crispy segmentation is obtained at the root level. This approach is applied to segment a variety of composite texture images into homogeneous texture areas and very good segmentation results are reported. 1

The interpretation of medical evidence is normally presented in terms of a controlled, but diversely expressed specialist vocabulary and natural language phrases. Such informally expressed data require human intervention to ascertain its relevance in any specific case. In order to facilitate machine-based reasoning about the evidence gathered, additional interpretive semantics must be attached to the data; a shift from a merely data-intensive approach to a semantics-rich model of evidence. In this paper, we present a system to formally annotate medical images captured to aid the diagnosis and management of breast cancer, that enables a series of semantics-based operations to be performed. Our approach is grounded upon an imaging ontology specifying the domain knowledge and a description logic (DL) taxonomic inferential engine responsible for semantics-based reasoning and image retrieval.

Two recent approaches are combined in a new technique to find moving arbitrary shapes. We combine the Velocity Hough Transform, which extracts moving conic sections, with a continuous formulation for arbitrary shape extraction, which avoids discretisation errors associated with GHT methods. The new approach has been evaluated on synthetic and real imagery and is demonstrated to provide motion analysis that is resilient to noise and to be able to detect its target shapes, which are both moving and arbitrary. Further, it is shown to have performance advantages over contemporaneous single-image extraction techniques. Finally, it appears to offer improved immunity to noise and occlusion, consistent with evidence gathering techniques, as shown by results on real images. 1

No abstract available.

The extraction of curvilinear features from images is an important task in many image analysis applications. We describe an approach to such tasks based on the combination of an efficient optimal path finding algorithm and the use of window constrained search spaces. We conclude that the combination provides an effective vehicle for curvilinear feature extraction from a variety of images.>

Describes a method for locating and extracting ridge and valley (and hence line) features in grey level images. A representation of the image known as the extended facet region model (EFRM) is first formed. The primary grouping constraint in this process is the orientation of patches in the image intensity surface. Inspection of geometric relationships between entities in the EFRM leads to the identification of ridge and valley structures and the subsequent placement of extracted line paths. The structures have knowledge of the local topological form of the image.>

Proposes a colour distance measure based on the new TekHVC colour space and demonstrates its application to the problem of colour image segmentation. Two segmentation algorithms are presented and compared-relaxation at the boundaries of a split and merge and an optimal region grower. The authors conclude that the output of the region grower is the most suitable for the purposes of subsequent region grouping and object recognition.>

A method of modeling image surfaces that combines B.J. Burns' (1986) orientation prominence work with R.M. Haralick's (1980) facet modeling and extends these ideas to the surface region-manipulation level is introduced. The proposed intermediate representation for grey level images retains a maximal amount of implicit image information in a form geared to high-level processing, so that features can be manipulated without explicit placement of edges. The proposed model, known as the extended facet region model (EFRM), performs surface fitting, leading to a segmentation of the image where the gradient orientation of surface patches is the initial grouping parameter. The image representation scheme developed so far has been successfully applied to several problem sets within the vision field.>

This paper describes the control of a large industrial robot with a unique mechanical structure. A triangular truss structure is employed that provides high lifting force capability while minimizing flexure due to loading. Homogeneous transformations are developed which allow control of the robot using Cartesian coordinates. Although the robot is not of the typical serial link configuration, a geometric approach can be used that results in a transformation matrix of standard form.