After decades of unrealistic predictions and expectations, robots have finally escaped from industrial workplaces and made their way into our homes,offices, museums and other public spaces. These service robots are increasingly present in our environments and many believe that it is in the area ofservice and domestic robotics that we will see the largest growth within thenext few years. In order to realize the dream of robot assistants performing human-like tasks together with humans in a seamless fashion, we need toprovide them with the fundamental capability of understanding complex, dynamic and unstructured environments. More importantly, we need to enablethem the sharing of our understanding of space to permit natural cooper-ation. To this end, this thesis addresses the problem of building internalrepresentations of space for artificial mobile agents populated with humanspatial semantics as well as means for inferring that semantics from sensoryinformation. More specifically, an extensible approach to place classificationis introduced and used for mobile robot localization as well as categorizationand extraction of spatial semantic concepts from general place appearance andgeometry. The models can be incrementally adapted to the dynamic changesin the environment and employ efficient ways for cue integration, sensor fu-sion and confidence estimation. In addition, a system and representationalapproach to semantic mapping is presented. The system incorporates and in-tegrates semantic knowledge from multiple sources such as the geometry andgeneral appearance of places, presence of objects, topology of the environmentas well as human input. A conceptual map is designed and used for modelingand reasoning about spatial concepts and their relations to spatial entitiesand their semantic properties. Finally, the semantic mapping algorithm isbuilt into an integrated robotic system and shown to substantially enhancethe performance of the robot on the complex task of active object search. Thepresented evaluations show the effectiveness of the system and its underlyingcomponents and demonstrate applicability to real-world problems in realistichuman settings. / QC 20110527

This thesis addresses the problem of people detection and tracking by mobile robots in indoor environments. A system that can detect and recognise people is an essential part of any mobile robot that is designed to operate in populated environments. Information about the presence and location of persons in the robot’s surroundings is necessary to enable interaction with the human operator, and also for ensuring the safety of people near the robot. The presented people tracking system uses a combination of thermal and colour information to robustly track persons. The use of a thermal camera simplifies the detection problem, which is especially difficult on a mobile platform. The system is based on a fast and efficient samplebased tracking method that enables tracking of people in real-time. The elliptic measurement model is fast to calculate and allows detection and tracking of persons under different views. An explicit model of the human silhouette effectively distinguishes persons from other objects in the scene. Moreover the process of detection and localisation is performed simultaneously so that measurements are incorporated directly into the tracking framework without thresholding of observations. With this approach persons can be detected independently from current light conditions and in situations where other popular detection methods based on skin colour would fail. A very challenging situation for a tracking system occurs when multiple persons are present on the scene. The tracking system has to estimate the number and position of all persons in the vicinity of the robot. Tracking of multiple persons in the presented system is realised by an efficient algorithm that mitigates the problems of combinatorial explosion common to other known algorithms. A sequential detector initialises an independent tracking filter for each new person appearing in the image. A single filter is automatically deleted when it stops tracking a person. While thermal vision is good for detecting people, it can be very difficult to maintain the correct association between different observations and persons, especially where they occlude one another, due to the unpredictable appearance and social behaviour of humans. To address these problems the presented tracking system uses additional information from the colour camera. An adaptive colour model is incorporated into the measurement model of the tracker to improve data association. For this purpose an efficient integral image based method is used to maintain the real-time performance of the tracker. To deal with occlusions the system uses an explicit method that first detects situations where people occlude each other. This is realised by a new approach based on a machine learning classifier for pairwise comparison of persons that uses both thermal and colour features provided by the tracker. This information is then incorporated into the tracker for occlusion handling and to resolve situations where persons reappear in a scene. Finally the thesis presents a comprehensive, quantitative evaluation of the whole system and its different components using a set of well defined performance measures. The behaviour of the system was investigated on different data sets including different real office environments and different appearances and behaviours of persons. Moreover the influence of all important system parameters on the performance of the system was checked and their values optimised based on these results.

Embedded network software has become increasingly interesting for both research and business as more and more networked embedded systems emerge. Well-known infrastructure protocol stacks are reimplemented on new embedded hardware and software architectures. New requirements of modern applications and devices require to implement newly designed or revised protocols. However, implementing protocol stacks for embedded systems remains a time-consuming and error-prone task due to the complexity and performancecritical nature of network software. It is even more so when targeting resource constrained embedded systems: implementations have to minimize energy consumption, memory usage etc., while programming efficiency is needed to improve on time-to-market, scalability, maintainability and product evolution. Therefore, it is worth researching on how to make protocol stack implementations for embedded systems both easier and more likely to be correct within the resource limits. In the work presented in this thesis, we take a language-based approach and aim to facilitate the implementation of protocol stacks while realizing performance demands and being aware of energy consumption and memory usage within the constraints imposed by embedded systems. We give background on DSL implementation techniques, investigate common practices in network protocol development to determine the potential of domain-specifi languages (DSLs) for embedded network software, and propose a domain-specifi embedded language (DSEL), Protege (Protocol Implementation Generator), for declaratively describing overlaid protocol stacks. In Protege, a high-level packet specification is dually compiled into an internal data representation for protocol logic implementation, and packet processing methods which are then integrated into the dataflow framework of a protocol overlay specification. Constructs for finite state machines allow to specify protocol logic in a concise manner, close to the protocol specification style. Protege specifications are compiled to highly portable C code for various architectures. Four attached scientific papers report our main results in more detail: an embedded implementation of the data description calculus in Haskell, a compilation framework for generating packet processing code with overlays, the domain-specific language Protege in overview (including embedding techniques and runtime system features), and a real-world case study implementing an industrial application protocol.

The mobile market is growing. More and more business units, (BUs) at the multinational company ABB start considering utilizing mobility to create business value, by offering product/system enhancement and by offering new services. However, at present, there is no standard way to realize this and they struggle with the security challenges for their solution. This thesis work will focus on investigating the Mobile Device Management (MDM) market and analyze how the existing technologies address all security requirements of a mobility solution. The main goal is to investigate how multi criteria methods can be applied to facilitate the choice of the most adequate MDM tool. It needs to provide a secure common architecture and integrate with existing backend infrastructures. The benefits include: (1) enabling the employees to use their own mobile devices in a secure way for working with the enterprise, (2) reliably integrating a backend system for new services in enterprise mobility, and (3) providing a (proven) secure solution to control and manage mobility on MDM technology.

A video database contains an enormous amount of information. In order to search through the database, metadata can be attached to each video. One such type of metadata can be labels containing speakers and where they are speaking. With the help of speaker recognition this type of metadata can automatically be assigned to each video. In this thesis a speaker recognition plug-in for Vidispine, an API media asset management platform, is presented. The plug-in was developed with the help of the LIUM SpkDiarization toolkit for speaker diarization and the ALIZE/LIA RAL toolkit for speaker identification. The choice of using the method of GMM-UBM that ALIZE/LIA RAL offers, was made through an in-depth theoretical study of different identification methods. The in-depth study is presented in its own chapter. The goal of the plug-in was to perform an identification rate of 85%. However, the results unfortunately became as low as 63%. Among the issues the plug-in faces, its low performance on female speaker was shown to be crucial.

This thesis is a comprehensive study of matrix factorization methods used in recommender systems. We study and analyze the existing models, specifically probabilistic models used in conjunction with matrix factorization methods, for recommender systems from a machine learning perspective. We implement two different methods suggested in scientific literature and conduct experiments on the prediction accuracy of the models on the Yahoo! Movies rating dataset.

Diss. (sammanfattning) Stockholm : Univ., 2005. / Härtill 6 uppsatser.

<p>Målet med detta arbete är att undersöka hur händelsedetektering i aktiva databassystem kan visualiseras samt identifiera vilka faktorer som har en central roll vid denna visualisering.</p><p>Arbetet jämför olika tekniker som kan användas för att visualisera händelsedetektering med hänseende till några olika faktorer. Dessa faktorer har identifierats ur litteratur om dels händelsedetektering och dels visualisering av regler.</p><p>Resultatet visar vilka faktorer som har stor påverkan vid visualisering av händelsedetektering samt hur de olika teknikerna uppfyller dessa faktorer. Vidare ges ett förslag på hur olika tekniker kan användas tillsammans för att skapa ett gränssnitt till ett program som förklarar consumption modes.</p>

<p>Denna rapport beskriver varför transmission av data över telenätet är problematiskt och hur datakomprimering kan användas för att öka hastigheten på transmissionen. Datakomprimering används redan nu för att hålla nere storleken på den data som ska skickas, t ex är ofta bilder i JPEG-format eller något annat komprimerat bildformat. Rapporten inriktar sig på komprimering av HTML-dokument då dessa i nuläget inte komprimeras före transmissionen. Först beskrivs ett antal vanliga komprimeringsalgoritmer och sedan väljs en metod ut som verkar lämplig för komprimering av HTML-dokument. Algoritmen modifieras till att utföra bättre komprimering på HTML-dokument genom att använda en kombination av förlustkomprimering och "överblivna" ASCII-koder. Algoritmen implementeras och testas med avseende på komprimeringseffektivitet och komprimeringshastighet.</p><p>All data komprimeras av modemet innan den skickas iväg, men hypotesen i denna rapport är att en implementerad algoritm som exekveras på processorn på den dator modemet sitter kopplat till komprimerar mera effektivt. Därför jämförs algoritmen med MNP-5 och V.42bis protokollen som används för datakomprimering i moderna modem. Då algoritmen tycks komprimera bättre än båda de protokoll som nämns ovan, bör den kunna användas som komplement till komprimeringen i modemen. Den kan dock inte helt ersätta komprimeringen i modemen då den enda komprimering som kan utföras av algoritmen är den av HTML-dokument.</p>

<p>I denna rapport behandlas frågor som rör extern validering och datorstöd (CASE) för detta. Rapporten redogör för en undersökning vars syfte är att ta reda på vilka krav som tekniker för extern validering ställer på CASE-verktyg och hur verktygen svarar upp till dessa krav. Olika metoder finns för att skapa ett informationssystem och varje metod har sina beskrivningstekniker som underlag för extern validering.</p><p>De beskrivningstekniker som föreskrivs i en metod är ofta mycket detaljerade. En metods notationen ställer därför stora krav på CASE-verktyg som uttalar stöd för denna metod.</p><p>Arbetet som beskrivs i denna rapport identifierar krav på datorstöd utifrån en viss metods beskrivningstekniker samt ytterligare valideringstekniker som beskrivs i litteraturen. Därefter undersöks CASE-verktygs förmåga att svara upp till dessa krav.</p><p>Själva undersökningen av verktygen genomförs via ett testförfarande där samtliga utvalda verktyg implementerar två olika testfall.</p><p>Resultatet från genomförandet av tester visar att det existerar stora skillnader mellan verktygens påstådda stöd för en viss metod och vad de faktiskt klarar av. CASE-verktygens förmåga att ge stöd för extern validering enligt vald metod är överraskande dåligt. Stöd för andra valideringstekniker än de som föreskrivs enligt vald metod är i stort sett obefintligt.</p>