Towards anomaly detection for increased security in multibiometric systems: spoofing-resistant 1-median fusion eliminating outliers

Multibiometrics aims at improving biometric security in presence of spoofing attempts, but exposes a larger availability of points of attack. Standard fusion rules have been shown to be highly sensitive to spoofing attempts – even in case of a single fake instance only. This paper presents a novel spoofing-resistant fusion scheme proposing the detection and elimination of anomalous fusion input in an ensemble of evidence with liveness information. This approach aims at making multibiometric systems more resistant to presentation attacks by modeling the typical behaviour of human surveillance operators detecting anomalies as employed in many decision support systems. It is shown to improve security, while retaining the high accuracy level of standard fusion approaches on the latest Fingerprint Liveness Detection Competition (LivDet) 2013 dataset.

Infrared image enhancement based on hybrid-domain consideration and data fusion methods

Au cours des dernières décennies, l’effort sur les applications de capteurs infrarouges a largement progressé dans le monde. Mais, une certaine difficulté demeure, en ce qui concerne le fait que les objets ne sont pas assez clairs ou ne peuvent pas toujours être distingués facilement dans l’image obtenue pour la scène observée. L’amélioration de l’image infrarouge a joué un rôle important dans le développement de technologies de la vision infrarouge de l’ordinateur, le traitement de l’image et les essais non destructifs, etc. Cette thèse traite de la question des techniques d’amélioration de l’image infrarouge en deux aspects, y compris le traitement d’une seule image infrarouge dans le domaine hybride espacefréquence, et la fusion d’images infrarouges et visibles employant la technique du nonsubsampled Contourlet transformer (NSCT). La fusion d’images peut être considérée comme étant la poursuite de l’exploration du modèle d’amélioration de l’image unique infrarouge, alors qu’il combine les images infrarouges et visibles en une seule image pour représenter et améliorer toutes les informations utiles et les caractéristiques des images sources, car une seule image ne pouvait contenir tous les renseignements pertinents ou disponibles en raison de restrictions découlant de tout capteur unique de l’imagerie. Nous examinons et faisons une enquête concernant le développement de techniques d’amélioration d’images infrarouges, et ensuite nous nous consacrons à l’amélioration de l’image unique infrarouge, et nous proposons un schéma d’amélioration de domaine hybride avec une méthode d’évaluation floue de seuil amélioré, qui permet d’obtenir une qualité d’image supérieure et améliore la perception visuelle humaine. Les techniques de fusion d’images infrarouges et visibles sont établies à l’aide de la mise en oeuvre d’une mise en registre précise des images sources acquises par différents capteurs. L’algorithme SURF-RANSAC est appliqué pour la mise en registre tout au long des travaux de recherche, ce qui conduit à des images mises en registre de façon très précise et des bénéfices accrus pour le traitement de fusion. Pour les questions de fusion d’images infrarouges et visibles, une série d’approches avancées et efficaces sont proposés. Une méthode standard de fusion à base de NSCT multi-canal est présente comme référence pour les approches de fusion proposées suivantes. Une approche conjointe de fusion, impliquant l’Adaptive-Gaussian NSCT et la transformée en ondelettes (Wavelet Transform, WT) est propose, ce qui conduit à des résultats de fusion qui sont meilleurs que ceux obtenus avec les méthodes non-adaptatives générales. Une approche de fusion basée sur le NSCT employant la détection comprime (CS, compressed sensing) et de la variation totale (TV) à des coefficients d’échantillons clairsemés et effectuant la reconstruction de coefficients fusionnés de façon précise est proposée, qui obtient de bien meilleurs résultats de fusion par le biais d’une pré-amélioration de l’image infrarouge et en diminuant les informations redondantes des coefficients de fusion. Une procédure de fusion basée sur le NSCT utilisant une technique de détection rapide de rétrécissement itératif comprimé (fast iterative-shrinking compressed sensing, FISCS) est proposée pour compresser les coefficients décomposés et reconstruire les coefficients fusionnés dans le processus de fusion, qui conduit à de meilleurs résultats plus rapidement et d’une manière efficace.

Regionen och EU? : Uppfattningar och attityder till EU-relaterade frågor i Västsverige

This thesis explores perceptions and preferences on regional action in EU-related frameworks among regional actors in Western Sweden. Building upon the literature on Europeanisation and the Fusion approach, three dimensions of Europeanisation are clarified and explored– download, upload and crossload – and together with a set of five variables that constitute the Micro Fusion Framework; a comprehensive analytical tool is developed. The thesis analyses the intense debate among the members of West Sweden that took place from 2011 to 2013 that focused on how to functionally organise the regional office in Brussels in order to meet future challenges. Surprisingly, the members eventually decided to terminate their cooperation and close the jointly owned office in Brussels in spite of the fact that it has been widely regarded as successful and effective. Diverging perceptions and preferences is understood in terms of three positions on regional action; a download-, upload- and a coherent oriented position. Finally, the thesis presents the empirical findings and discusses in relation to three fusion scenarios, infusion, defusion and clustered fusion. In terms of Micro Fusion Framework, the dynamics shaping why West Sweden was finally regarded as a dysfunctional arena for regional action are explained by a shift of attention and action among regional actors in Western Sweden that led to pressure for further institutional adaptation in order to meet the demand of how ‘to get the best out of the EU’. Further, this redefinition of how to handle EU-affairs within the upload-oriented position was accompanied by positive attitudes towards the potential to bypass the state and thereby pursue regional priorities directly in Brussels given the compound nature of the EU. In contrast, those regional actors that are found to be more download-oriented often question the benefits of uploading activities in practice and advocate close relations to the state. A coherent oriented position recognises the importance of activities related to both of the vertical dimensions of Europeanisation.

Robot visual localization through local feature fusion: an evaluation of multiple classifiers combination approaches

In the last decade, local image features have been widely used in robot visual localization. In order to assess image similarity, a strategy exploiting these features compares raw descriptors extracted from the current image with those in the models of places. This paper addresses the ensuing step in this process, where a combining function must be used to aggregate results and assign each place a score. Casting the problem in the multiple classifier systems framework, in this paper we compare several candidate combiners with respect to their performance in the visual localization task. For this evaluation, we selected the most popular methods in the class of non-trained combiners, namely the sum rule and product rule. A deeper insight into the potential of these combiners is provided through a discriminativity analysis involving the algebraic rules and two extensions of these methods: the threshold, as well as the weighted modifications. In addition, a voting method, previously used in robot visual localization, is assessed. Furthermore, we address the process of constructing a model of the environment by describing how the model granularity impacts upon performance. All combiners are tested on a visual localization task, carried out on a public dataset. It is experimentally demonstrated that the sum rule extensions globally achieve the best performance, confirming the general agreement on the robustness of this rule in other classification problems. The voting method, whilst competitive with the product rule in its standard form, is shown to be outperformed by its modified versions.

Unveiling the mating system and genetic variability in the yeast Kwoniella mangroviensis using molecular approaches

Dissertação para obtenção do Grau de Mestre em Genética Molecular e Biomedicina

Data fusion with computational intelligence techniques: a case study of fuzzy inference for terrain assessment

Dissertação apresentada para obtenção do Grau de Mestre em Engenharia Electrotécnica e de Computadores, pela Universidade Nova de Lisboa, Faculdade de Ciências e Tecnologia

KIAA1549: BRAF gene fusion and FGFR1 hotspot mutations are prognostic factors in pilocytic astrocytomas

Up to 20% of patients with pilocytic astrocytoma (PA) experience a poor outcome. BRAF alterations and Fibroblast growth factor receptor 1 (FGFR1) point mutations are key molecular alterations in Pas, but their clinical implications are not established. We aimed to determine the frequency and prognostic role of these alterations in a cohort of 69 patients with PAs. We assessed KIAA1549:BRAF fusion by fluorescence in situ hybridization and BRAF (exon 15) mutations by capillary sequencing. In addition, FGFR1 expression was analyzed using immunohistochemistry, and this was compared with gene amplification and hotspot mutations (exons 12 and 14) assessed by fluorescence in situ hybridization and capillary sequencing. KIAA1549:BRAF fusion was identified in almost 60% of cases. Two tumors harbored mutated BRAF. Despite high FGFR1 expression overall, no cases had FGFR1 amplifications. Three cases harbored a FGFR1 p.K656E point mutation. No correlation was observed between BRAF and FGFR1 alterations. The cases were predominantly pediatric (87%), and no statistical differences were observed in molecular alterations-related patient ages. In summary, we confirmed the high frequency of KIAA1549:BRAF fusion in PAs and its association with a better outcome. Oncogenic mutations of FGFR1, although rare, occurred in a subset of patients with worse outcome. These molecular alterations may constitute alternative targets for novel clinical approaches, when radical surgical resection is unachievable.

New integrative computational approaches unveil the Saccharomyces cerevisiae pheno-metabolomic fermentative profile and allow strain selection for winemaking

During must fermentation by Saccharomyces cerevisiae strains thousands of volatile aroma compounds are formed. The objective of the present work was to adapt computational approaches to analyze pheno-metabolomic diversity of a S. cerevisiae strain collection with different origins. Phenotypic and genetic characterization together with individual must fermentations were performed, and metabolites relevant to aromatic profiles were determined. Experimental results were projected onto a common coordinates system, revealing 17 statistical-relevant multi-dimensional modules, combining sets of most-correlated features of noteworthy biological importance. The present method allowed, as a breakthrough, to combine genetic, phenotypic and metabolomic data, which has not been possible so far due to difficulties in comparing different types of data. Therefore, the proposed computational approach revealed as successful to shed light into the holistic characterization of S. cerevisiae pheno-metabolome in must fermentative conditions. This will allow the identification of combined relevant features with application in selection of good winemaking strains.

Novel approaches in anti-arenaviral drug development.

Hemorrhagic fevers caused by arenaviruses are among the most devastating emerging human diseases. Considering the number of individuals affected, the current lack of a licensed vaccine, and the limited therapeutic options, arenaviruses are arguably among the most neglected tropical pathogens and the development of efficacious anti-arenaviral drugs is of high priority. Over the past years significant efforts have been undertaken to identify novel potent inhibitors of arenavirus infection. High throughput screening of small molecule libraries employing pseudotype platforms led to the discovery of several potent and broadly active inhibitors of arenavirus cell entry that are effective against the major hemorrhagic arenaviruses. Mechanistic studies revealed that these novel entry inhibitors block arenavirus membrane fusion and provided novel insights into the unusual mechanism of this process. The success of these approaches highlights the power of small molecule screens in antiviral drug discovery and establishes arenavirus membrane fusion as a robust drug target. These broad screenings have been complemented by strategies targeting cellular factors involved in productive arenavirus infection. Approaches targeting the cellular protease implicated in maturation of the fusion-active viral envelope glycoprotein identified the proteolytic processing of the arenavirus glycoprotein precursor as a novel and promising target for anti-arenaviral strategies.

Multi-modal Change Detection, Application to the Detection of Flooded Areas: Outcome of the 2009-2010 data fusion contest

The 2009-2010 Data Fusion Contest organized by the Data Fusion Technical Committee of the IEEE Geoscience and Remote Sensing Society was focused on the detection of flooded areas using multi-temporal and multi-modal images. Both high spatial resolution optical and synthetic aperture radar data were provided. The goal was not only to identify the best algorithms (in terms of accuracy), but also to investigate the further improvement derived from decision fusion. This paper presents the four awarded algorithms and the conclusions of the contest, investigating both supervised and unsupervised methods and the use of multi-modal data for flood detection. Interestingly, a simple unsupervised change detection method provided similar accuracy as supervised approaches, and a digital elevation model-based predictive method yielded a comparable projected change detection map without using post-event data.

Model-based Segmentation and Fusion of 3D Computed Tomography and 3D Ultrasound of the Eye for Radiotherapy Planning

Computed Tomography (CT) represents the standard imaging modality for tumor volume delineation for radiotherapy treatment planning of retinoblastoma despite some inherent limitations. CT scan is very useful in providing information on physical density for dose calculation and morphological volumetric information but presents a low sensitivity in assessing the tumor viability. On the other hand, 3D ultrasound (US) allows a highly accurate definition of the tumor volume thanks to its high spatial resolution but it is not currently integrated in the treatment planning but used only for diagnosis and follow-up. Our ultimate goal is an automatic segmentation of gross tumor volume (GTV) in the 3D US, the segmentation of the organs at risk (OAR) in the CT and the registration of both modalities. In this paper, we present some preliminary results in this direction. We present 3D active contour-based segmentation of the eye ball and the lens in CT images; the presented approach incorporates the prior knowledge of the anatomy by using a 3D geometrical eye model. The automated segmentation results are validated by comparing with manual segmentations. Then, we present two approaches for the fusion of 3D CT and US images: (i) landmark-based transformation, and (ii) object-based transformation that makes use of eye ball contour information on CT and US images.

Model-based Segmentation and Image Fusion of 3D Computed Tomography and 3D Ultrasound of the Eye for Radiotherapy Planning

For radiotherapy treatment planning of retinoblastoma inchildhood, Computed Tomography (CT) represents thestandard method for tumor volume delineation, despitesome inherent limitations. CT scan is very useful inproviding information on physical density for dosecalculation and morphological volumetric information butpresents a low sensitivity in assessing the tumorviability. On the other hand, 3D ultrasound (US) allows ahigh accurate definition of the tumor volume thanks toits high spatial resolution but it is not currentlyintegrated in the treatment planning but used only fordiagnosis and follow-up. Our ultimate goal is anautomatic segmentation of gross tumor volume (GTV) in the3D US, the segmentation of the organs at risk (OAR) inthe CT and the registration of both. In this paper, wepresent some preliminary results in this direction. Wepresent 3D active contour-based segmentation of the eyeball and the lens in CT images; the presented approachincorporates the prior knowledge of the anatomy by usinga 3D geometrical eye model. The automated segmentationresults are validated by comparing with manualsegmentations. Then, for the fusion of 3D CT and USimages, we present two approaches: (i) landmark-basedtransformation, and (ii) object-based transformation thatmakes use of eye ball contour information on CT and USimages.

Detection of Clinically Significant Prostate Cancer Using Magnetic Resonance Imaging-Ultrasound Fusion Targeted Biopsy: A Systematic Review.

CONTEXT: The current standard for diagnosing prostate cancer in men at risk relies on a transrectal ultrasound-guided biopsy test that is blind to the location of the cancer. To increase the accuracy of this diagnostic pathway, a software-based magnetic resonance imaging-ultrasound (MRI-US) fusion targeted biopsy approach has been proposed. OBJECTIVE: Our main objective was to compare the detection rate of clinically significant prostate cancer with software-based MRI-US fusion targeted biopsy against standard biopsy. The two strategies were also compared in terms of detection of all cancers, sampling utility and efficiency, and rate of serious adverse events. The outcomes of different targeted approaches were also compared. EVIDENCE ACQUISITION: We performed a systematic review of PubMed/Medline, Embase (via Ovid), and Cochrane Review databases in December 2013 following the Preferred Reported Items for Systematic reviews and Meta-analysis statement. The risk of bias was evaluated using the Quality Assessment of Diagnostic Accuracy Studies-2 tool. EVIDENCE SYNTHESIS: Fourteen papers reporting the outcomes of 15 studies (n=2293; range: 13-582) were included. We found that MRI-US fusion targeted biopsies detect more clinically significant cancers (median: 33.3% vs 23.6%; range: 13.2-50% vs 4.8-52%) using fewer cores (median: 9.2 vs 37.1) compared with standard biopsy techniques, respectively. Some studies showed a lower detection rate of all cancer (median: 50.5% vs 43.4%; range: 23.7-82.1% vs 14.3-59%). MRI-US fusion targeted biopsy was able to detect some clinically significant cancers that would have been missed by using only standard biopsy (median: 9.1%; range: 5-16.2%). It was not possible to determine which of the two biopsy approaches led most to serious adverse events because standard and targeted biopsies were performed in the same session. Software-based MRI-US fusion targeted biopsy detected more clinically significant disease than visual targeted biopsy in the only study reporting on this outcome (20.3% vs 15.1%). CONCLUSIONS: Software-based MRI-US fusion targeted biopsy seems to detect more clinically significant cancers deploying fewer cores than standard biopsy. Because there was significant study heterogeneity in patient inclusion, definition of significant cancer, and the protocol used to conduct the standard biopsy, these findings need to be confirmed by further large multicentre validating studies. PATIENT SUMMARY: We compared the ability of standard biopsy to diagnose prostate cancer against a novel approach using software to overlay the images from magnetic resonance imaging and ultrasound to guide biopsies towards the suspicious areas of the prostate. We found consistent findings showing the superiority of this novel targeted approach, although further high-quality evidence is needed to change current practice.

Optimising approaches to active immunotherapy of cancer

The design of therapeutic cancer vaccines is aimed at inducing high numbers and potent T cells that are able to target and eradicate malignant cells. This calls for close collaboration between cells of the innate immune system, in particular dendritic cells (DCs), and cells of the adaptive immune system, notably CD4+ helper T cells and CD8+ cytotoxic T cells. Therapeutic vaccines are aided by adjuvants, which can be, for example, Toll¬like Receptor agonists or agents promoting the cytosolic delivery of antigens, among others. Vaccination with long synthetic peptides (LSPs) is a promising strategy, as the requirement for their intracellular processing will mainly target LSPs to professional antigen presenting cells (APCs), hence avoiding the immune tolerance elicited by the presentation of antigens by non-professional APCs. The unique property of antigen cross-processing and cross-presentation activity by DCs plays an important role in eliciting antitumour immunity given that antigens from engulfed dead tumour cells require this distinct biological process to be processed and presented to CD8+T cells in the context of MHC class I molecules. DCs expressing the XCR1 chemokine receptor are characterised by their superior capability of antigen cross- presentation and priming of highly cytotoxic T lymphocyte (CTL) responses. Recently, XCR1 was found to be also expressed in tissue-residents DCs in humans, with a simitar transcriptional profile to that of cross- presenting murine DCs. This shed light into the value of harnessing this subtype of XCR1+ cross-presenting DCs for therapeutic vaccination of cancer. In this study, we explored ways of adjuvanting and optimising LSP therapeutic vaccinations by the use, in Part I, of the XCLl chemokine that selectively binds to the XCR1 receptor, as a mean to target antigen to the cross-presenting XCR1+ DCs; and in Part II, by the inclusion of Q.S21 in the LSP vaccine formulation, a saponin with adjuvant activity, as well as the ability to promote cytosolic delivery of LSP antigens due to its intrinsic cell membrane insertion activity. In Part I, we designed and produced XCLl-(OVA LSP)-Fc fusion proteins, and showed that their binding to XCR1+ DCs mediate their chemoattraction. In addition, therapeutic vaccinations adjuvanted with XCLl-(OVA LSP)-Fc fusion proteins significantly enhanced the OVA-specific CD8+ T cell response, and led to complete tumour regression in the EL4-OVA model, and significant control of tumour growth in the B16.0VA tumour model. With the aim to optimise the co-delivery of LSP antigen and XCLl to skin-draining lymph nodes we also tested immunisations using nanoparticle (NP)-conjugated OVA LSP in the presence or absence of XCLl chemokine. The NP-mediated delivery of LSP potentiated the CTL response seen in the blood of vaccinated mice, and NP-OVA LSP vaccine in the presence of XCLl led to higher blood frequencies of OVA-specific memory-precursor effector cells. Nevertheless, in these settings, the addition XCLl to NP-OVA LSP vaccine formulation did not increase its antitumour therapeutic effect. In the Part II, we assessed in HLA-A2/DR1 mice the immunogenicity of the Melan-AA27L LSP or the Melan-A26. 35 AA27l short synthetic peptide (SSP) used in conjunction with the saponin adjuvant QS21, aiming to identify a potent adjuvant formulation that elicits a quantitatively and qualitatively strong immune response to tumour antigens. We showed a high CTL immune response elicited by the use of Melan-A LSP or SSP with QS21, which both exerted similar killing capacity upon in vivo transfer of target cells expressing the Melan-A peptide in the context of HLA-A2 molecules. However, the response generated by the LSP immunisation comprised higher percentages of CD8+T cells of the central memory phenotype (CD44hl CD62L+ and CCR7+ CD62L+) than those of SSP immunisation, and most importantly, the strong LSP+QS21 response was strictly CD4+T cell-dependent, as shown upon CD4 T cell depletion. Altogether, these results suggest that both XCLl and QS21 may enhance the ability of LSP to prime CD8 specific T cell responses, and promote a long-term memory response. Therefore, these observations may have important implications for the design of protein or LSP-based cancer vaccines for specific immunotherapy of cancer -- Les vacans thérapeutiques contre le cancer visent à induire une forte et durable réponse immunitaire contre des cellules cancéreuses résiduelles. Cette réponse requiert la collaboration entre le système immunitaire inné, en particulier les cellules dendrites (DCs), et le système immunitaire adaptatif, en l'occurrence les lymphocytes TCD4 hdper et CD8 cytotoxiques. La mise au point d'adjuvants et de molécules mimant un agent pathogène tels les ligands TLRs ou d'autres agents facilitant l'internalisation d'antigènes, est essentielle pour casser la tolérance du système immunitaire contre les cellules cancéreuses afin de générer une réponse effectrice et mémoire contre la tumeur. L'utilisation de longs peptides synthétiques (LSPs) est une approche prometteuse du fait que leur présentation en tant qu'antigénes requiert leur internalisation et leur transformation par les cellules dendrites (DCs, qui sont les mieux à même d'éviter la tolérance immunitaire. Récemment une sous-population de DCs exprimant le récepteur XCR1 a été décrite comme ayant une capacité supérieure dans la cross-présentation d'antigènes, d'où un intérêt à développer des vaccins ciblant les DCs exprimant le XCR1. Durant ma thèse de doctorat, j'ai exploré différentes approches pour optimiser les vaccins avec LSPs. La première partie visait à cibler les XCR1-DCs à l'aide de la chemokine XCL1 spécifique du récepteur XCR1, soit sou s la forme de protéine de fusion XCL1-OVA LSP-Fc, soit associée à des nanoparticules. La deuxième partie a consisté à tester l'association des LSPs avec I adjuvant QS21 dérivant d'une saponine dans le but d'optimiser l'internalisation cytosolique des longs peptides. Les protéines de fusion XCLl-OVA-Fc développées dans la première partie de mon travail, ont démontré leur capacité de liaison spécifique sur les XCRl-DCs associée à leur capacité de chemo-attractio. Lorsque inclues dans une mmunisation de souris porteuse de tumeurs établies, ces protéines de fusion XCL1-0VA LSP-Fc et XCLl-Fc plus OVA LSP ont induites une forte réponse CDS OVA spécifique permettant la complète régression des tumeurs de modèle EL4- 0VA et un retard de croissance significatif de tumeurs de type B16-0VA. Dans le but d'optimiser le drainage des LSPs vers es noyaux lymphatiques, nous avons également testé les LSPs fixés de manière covalente à des nanoparticules co- injectees ou non avec la chemokine XCL1. Cette formulation a également permis une forte réponse CD8 accompagnée d'un effet thérapeutique significatif, mais l'addition de la chemokine XCL1 n'a pas ajouté d'effet anti-tumeur supplémentaire. Dans la deuxième partie de ma thèse, j'ai comparé l'immunogénicité de l'antigène humain Melan A soit sous la forme d un LSP incluant un épitope CD4 et CD8 ou sous la forme d'un peptide ne contenant que l'épitope CD8 (SSP) Les peptides ont été formulés avec l'adjuvant QS21 et testés dans un modèle de souris transgéniques pour les MHC let II humains, respectivement le HLA-A2 et DR1. Les deux peptides LSP et SSP ont généré une forte réponse CD8 similaire assoc.ee a une capacité cytotoxique équivalente lors du transfert in vivo de cellules cibles présentant le peptide SSP' Cependant les souris immunisées avec le Melan A LSP présentaient un pourcentage plus élevé de CD8 ayant un Phénotype «centra, memory» (CD44h' CD62L+ and CCR7+ CD62L+) que les souris immunisées avec le SSP, même dix mois après I'immunisation. Par ailleurs, la réponse CD8 au Melan A LSP était strictement dépendante des lymphocytes CD4, contrairement à l'immunisation par le Melan A SSP qui n'était pas affectée. Dans l'ensemble ces résultats suggèrent que la chemokine XCL1 et l'adjuvant QS21 améliorent la réponse CD8 à un long peptide synthétique, favorisant ainsi le développement d'une réponse anti-tumeur mémoire durable. Ces observations pourraient être utiles au développement de nouveau vaccins thérapeutiques contre les tumeurs.

Information Exchange and Fusion in Dynamic and Heterogeneous Distributed Environments

Context awareness, dynamic reconfiguration at runtime and heterogeneity are key characteristics of future distributed systems, particularly in ubiquitous and mobile computing scenarios. The main contributions of this dissertation are theoretical as well as architectural concepts facilitating information exchange and fusion in heterogeneous and dynamic distributed environments. Our main focus is on bridging the heterogeneity issues and, at the same time, considering uncertain, imprecise and unreliable sensor information in information fusion and reasoning approaches. A domain ontology is used to establish a common vocabulary for the exchanged information. We thereby explicitly support different representations for the same kind of information and provide Inter-Representation Operations that convert between them. Special account is taken of the conversion of associated meta-data that express uncertainty and impreciseness. The Unscented Transformation, for example, is applied to propagate Gaussian normal distributions across highly non-linear Inter-Representation Operations. Uncertain sensor information is fused using the Dempster-Shafer Theory of Evidence as it allows explicit modelling of partial and complete ignorance. We also show how to incorporate the Dempster-Shafer Theory of Evidence into probabilistic reasoning schemes such as Hidden Markov Models in order to be able to consider the uncertainty of sensor information when deriving high-level information from low-level data. For all these concepts we provide architectural support as a guideline for developers of innovative information exchange and fusion infrastructures that are particularly targeted at heterogeneous dynamic environments. Two case studies serve as proof of concept. The first case study focuses on heterogeneous autonomous robots that have to spontaneously form a cooperative team in order to achieve a common goal. The second case study is concerned with an approach for user activity recognition which serves as baseline for a context-aware adaptive application. Both case studies demonstrate the viability and strengths of the proposed solution and emphasize that the Dempster-Shafer Theory of Evidence should be preferred to pure probability theory in applications involving non-linear Inter-Representation Operations.