Activity classification based on sensor data is a challenging task. Many studies have focused on two main methods to enable activity classification; namely sensor level classification and body-model level classification. This study aims to enable activity classification across sensor domains by considering an e-textile garment and provide the groundwork for transferring the e-textile garment to a vision-based classifier. The framework is comprised of three main components that enable the successful transfer of the body-worn system to the vision-based classifier. The inter-class confusion of the activity space is quantified to allow an ideal prediction of known class accuracy for varying levels of error within the system. Methods for quantifying sensor and garment level error are undertaken to identify challenges specific to a body-worn system. These methods are then used to inform decisions related to the classification accuracy and threshold of the classifier. Using activities from a vision-based system known to the classifier, a user study was conducted to generate an observed set of activities from the body-worn system. The results indicate that the vision-based classifier used is user-independent and can successfully handle classification across sensor domains. / Master of Science

No description available.

Det finns många studier på hur främmande arter påverkar en arts återinförsel, men det finns fåsom har gjorts om infödda arters påverkan. Atlantlax (Salmo salar) kunde en gång i tidenmigrera från Vänern i Sverige till Klarälven på den norska sidan. Det kan den inte idag, menidag talas det om att återinföra laxen till Klarälven i Norge så att den kan utföra hela sinmigrationsrutt igen. I den älven har laxen en gång i tiden koexisterat med sina konkurrenterharr (Thymallus thymallus), och öring (Salmo trutta) men idag är frågan hur denkonkurrensen mellan dessa tre arter kan komma att påverka återinförseln av lax. För att få enbättre bild av hur utgången för en potentiell återintroduktion kan se ut, så var syftet meddenna undersökning att ta reda på mot vilken art dessa tre arter främst riktade sinaaggressionsbeteenden mot. I akvarieförsök med juvenila individer av dessa tre arter harundersökningarna tagit plats och aggressiva beteenden filmades när arterna var i olikakombinationer. Resultatet visade att harr var den mest aktiva av de tre arterna och lax denminst aktiva, samt att lax blev dominerad utav både harr och öring. Resultatet i denna studietyder på att en återintroduktion av lax kan innehålla svårigheter, men i naturen är ett habitatmer komplext än i ett akvarium vilket kan ge ett annat resultat som kanske gör enåterintroduktion mer framgångsrik. / There are many studies about how the effects from exotic species could affect a reintroductionof a species, but there are only a few that have been made about the effects from nativespecies. The Atlantic salmon (Salmo salar) could historically speaking, migrate from lakeVänern in Sweden to the River Klarälven in Norway. Today it cannot do this, but there is anongoing discussion about reintroducing the salmon to the Norwegian side of River Klarälven,so it once again can swim along its entire migration route. In that river, the salmon used tocoexist with its competitors, grayling (Thymallus thymallus) and trout (Salmo trutta), buttoday the question is how competition between these three species could affect thereintroduction of salmon. To get a better understanding of the outcome of a possiblereintroduction, the purpose of this study was to ascertain which species these three speciesmainly directed their agonistic behavior towards. The study took place in aquariums withjuvenile individuals of the three species. Agonistic behavior was filmed when the specieswere together in different combinations. The results show that grayling is the most active ofthe three species, while salmon is the least active, and salmon was dominated by both graylingand trout. The results from this study indicates that a reintroduction of salmon might befraught with problems, but in nature a habitat is more complex than in an aquarium, whichmeans that in nature the results could be different, which might make a reintroduction moresuccessful.

The use of deep learning methods is increasing in medical image analysis, e.g., segmentation of organs in medical images. Deep learning methods are highly dependent on a large amount of training data, a common obstacle for medical image analysis. This master thesis proposes a class-agnostic loss function as a method to train on incomplete data. The project used CT images from 1587 breast cancer patients, with a variety of available segmentation masks for each patient. The class-agnostic loss function is given labels for each class for each sample, in this project, for each segmentation mask for each CT slice. The label tells the loss function if the mask is an actual mask or just a placeholder. If it is a placeholder, the comparison of the predicted mask and the placeholder will not contribute to the loss value. The results show that it was possible to use the class-agnostic loss function to train a segmentation model with eight output masks, with data that never had all eight masks present at the same time and gain approximately the same performance as single-mask models.

Time Series are used in healthcare, meteorology, and many other fields. Rigorous research has been done to develop distance measures and classifying algorithms for time series. When a time series is classified, one can ask what changes should be made to the time series to classify it differently. A time series with the appropriate changes that make the classifier classify the time series as a different class is known as a counterfactual explanation. There exist model-dependent methods for creating counterfactual explanations. However, there exists a lack in the literature of a model agnostic method for creating counterfactual explanations for Time Series. This study aims to answer the following research question. ” How does a model agnostic method for counterfactuals for time series perform in terms of cost and compactness compared to model dependent algorithms for counterfactuals for time series?” To answer the research question, a model agnostic method for creating counterfactuals for time series was created named Multi-Objective Counterfactuals For Time Series. The Evaluation of the Multi-Objective Counterfactual Explanation For Time Series performed better than the modeldependent algorithms in Compactness but worse in Cost.

Videos manipulated by machine learning have rapidly increased online in the past years. So called deepfakes can depict people who never participated in a video recording by transposing their faces onto others in it. This raises the concern of authenticity of media, which demand for higher performing detection methods in forensics. Introduction of AI detectors have been of interest, but is held back today by their lack of interpretability. The objective of this thesis was therefore to examine what the explainable AI method local interpretable model-agnostic explanations (LIME) could contribute to forensic investigations of deepfake video.  An evaluation was conducted where 3 multimedia forensics evaluated the contribution of visual explanations of classifications when investigating deepfake video frames. The estimated contribution was not significant yet answers showed that LIME may be used to indicate areas to start examine. LIME was however not considered to provide sufficient proof to why a frame was classified as `fake', and would if introduced be used as one of several methods in the process. Issues were apparent regarding the interpretability of the explanations, as well as LIME's ability to indicate features of manipulation with superpixels.

No description available.

It is natural to suppose that mental and physical properties are importantly distinct. Yet whatever this difference is, it has to be compatible with interaction between the mind and the body. Satisfaction of these desiderata leads to a paradox. If you make the mind strongly separate from the body, then there is the problem of bringing them together. If you unite them, then there is the problem of preserving their distinctiveness. It is the aim of this thesis to resolve the paradox. From the outset, it is assumed that the nature of interaction is most satisfactorily explained by an account of mental properties in monistic terms. For reasons for space, the arguments of Materialism are concentrated upon at the exposure of Idealism. Three strategies are examined, and found wanting. First, an instance of a non-reductive account provided by Davidson's 'Anomalous Monism'. Here, mental properties seem to be left with no role in influencing behaviour. Second, a review of reductionist accounts, ranging from Identity Theories to Representationalism. Criticism focuses upon the failure of reductionism to explain the connection between the function of a conscious state and its particular character. A Materialist treats mental states as if they were part of the physical universe. This implies that the nature of these states may be understood through scientific investigation, in the same manner as all other phenomena. The third strategy is to deny the above implication: that is, deny the assertion that, by existing, all aspects of an object are thereby knowable. The ideas of Colin McGinn are discussed as an example of this position. Since his arguments are equally suitable for non-Materialist purposes, they do not constitute an exclusively Materialist solution to the above paradox. This thesis offers an alternative way of pursuing the above strategy. It argues that the relation between mental states and our ways of understanding phenomena, is such that we should not expect our theories about the nature of 'mind' and the 'physical world' to employ the same terms. These properties appear distinct, not because they are different substances, but because they occupy different sides of the ‘process of understanding’ - ‘thing understood’ relationship. For convenience, this position is referred to as ‘Agnostic Materialism’. As interaction between the mind and the body is compatible with the mind having no influence upon our behaviour, it is incumbent upon the thesis to defend Materialism against the claim that mental properties are epiphenomenal. This is achieved by teasing out two ways in which such properties are considered inert: either because the workings of the mind are independent of the body; or because the mind’s processes are irrelevant to those of the body. The first claim is seen arise from the difficulty of seeing the mind as part of the physical world - a difficulty removed by the arguments in the previous paragraph. The second claim gains plausibility through a mistaken adherence to certain models of scientific explanation.

La pensée athée fait partie intégrante de la société française depuis maintenant plusieurs siècles. Si de nombreux ouvrages se sont essayés à en livrer une approche idéologique ou philosophique, peu d’études se sont pourtant réellement attachées, jusqu’à maintenant, à en analyser les composantes, les subtilités, les ressorts ou les acteurs dans une démarche purement sociologique. C’est là l’objectif qui a guidé cette thèse doctorale pour enfin tenter d’apprécier plus en détails la réalité du fait athée en France. Pour appréhender au mieux cette réalité et afin d’étudier une population méconnue car souvent discrète, le travail de recherche entrepris ici s’est volontairement concentré sur l’une des rares associations à être exclusivement composées d’athées déclarés. Bien que très petite en taille et totalement inconnue du grand public, l’« Union des Athées » est en effet la plus ancienne association de ce type au niveau national et constitue par là un terrain d’étude idéal pour qui souhaite s’approcher au plus près de la "parole" athée. Par l’étude fouillée de son histoire et l’analyse de sa conception idéologique, puis au moyen d’une enquête inédite réalisée auprès de ses membres, cette thèse propose des clés pour une compréhension plus actuelle de la pensée athée et permet d’aborder l’athéisme et la laïcité sous un angle nouveau, contribuant ainsi à poser les bases d’une autre réflexion sur les problématiques de l’actualité. En remettant en perspective les notions mêmes d’"athéisme" et de "Dieu", elle introduit par ailleurs un nouveau postulat : considérer l’athéisme ne peut s’envisager sans prendre en compte le contexte civilisationnel dans lequel il se vit. / The atheist’ thought has been an integral part of French society now since several centuries. Even though many writing works have tried to make an ideological and philosophical approach, very few studies have been set out to analyze its constituents, subtleties, or actors, dealing with purely sociological methodology. It is what this doctoral thesis led as its objective for the purpose of attempting to assess the reality of the atheist fact in France. In order to comprehend better this reality and to study this unrecognized population because of its unassuming nature, the research carried out here has been voluntarily concentrated on one of the rare associations composed exclusively by self-declared atheists. Although very small and totally unknown by the general public, the “Atheists Union” is indeed the oldest association of this kind on the national wide level, and thus forms an ideal field of study for who wishes to close to the nearest way of the atheists’ talks. By studying his detailed history and analyzing his ideological conception, and then, by way of an original survey carried out among his members, this thesis propose the key of comprehension of most actual atheists’ thoughts, and allow to approach the atheism and secularism from a new angle, contributing in thus way to put down the bases of an another reflection on the current issues. Putting in perspective the knowledge itself of “atheism” and “God”, it introduces a new postulate: considering the atheism will not be able to foresee without taking into account the civilizational context in which it lives.

Mobile phones, tablet computers, laptops, desktops, and large screen displays are increasingly available to individuals for information access, often simultaneously. Dominant content access protocols, such as HTTP/1.1, do not take advantage of this device multiplicity and support information access from single devices only. Changing devices means restarting an information session. Using devices in conjunction with each other poses several challenges, which include the presentation of content on devices with diverse form factors and propagation of the content changes across these devices. In this dissertation, I report on the design and implementation of MIDAS - architecture and a prototype system for multi-device presentations. I propose a framework, called 12C, for characterizing multi-device systems and evaluate MIDAS within this framework. MIDAS is designed as a middleware that can work with multiple client-server architectures, such as the Web and context-aware Trellis, a non-Web hypertext system. It presents information content simultaneously on devices with diverse characteristics without requiring sensor-enhanced environments. The system adapts content elements for optimal presentation on the target device while also striving to retain fidelity with the original form from a human perceptual perspective. MIDAS reconfigures its presentation in response to user actions, availability of devices, and environmental context, such as a user's location or the time of day. I conducted a pilot study that explored human perception of similarity when image attributes such as size and color depth are modified in the process of presenting images on different devices. The results indicated that users tend to prefer scaling of images to color-depth reduction but gray scaling of images is preferable to either modification. Not all images scale equally gracefully; those dominated by natural elements or manmade structures scale exceptionally well. Images that depict recognizable human faces or textual elements should be scaled only to an extent that these features retain their integrity. Attributes of the 12C framework describe aspects of multi-device systems that include infrastructure, presentation, interaction, interface, and security. Based on these criteria, MIDAS is a flexible infrastructure, which lends itself to several content distribution and interaction strategies by separating client- and server-side configuration.