In 2021, Intel released a new generation of consumer processors that had two different types of cores; they called them the efficiency-cores and the performance-cores, or for short, the E-cores and the P-cores. The release of these processors sparked interest in the potential impact if they were introduced into the enterprise market. Therefore tests were made on the I5-13600k on its power consumption and how it performs in virtual machines. Since the architecture is so new, there have not been any tests on the power usage in different configurations, which is a hot topic with today's power prices. The first question to be answered was: "How can one get the new architecture to function in virtual machines without mixing the various types of cores?" To solve that, the command “virsh edit” was used and then the name of the virtual machine to choose which threads to use for the virtual machine. The second question was: “how different is the power consumption between different configurations?” And to answer that, a power meter was used to measure the power consumption, which resulted in the Linux virtual machines drawing less power than the Windows 10 virtual machines. The third question was: “what is the performance of the processor in virtual machines?” To answer that, a benchmarking tool called Geekbench was used, which showed that the P-core only machine performed best as expected. The fourth and last question was: "What is the use case for having different types of cores?" From the tests, these processors shouldn’t be produced for the enterprise market, but small companies and enthusiasts can use them for servers when they need more cores

The primary aim of this master's thesis is to explore the differences between the methods usedby the Belgian and Swedish Police in their investigations of online child sexual abuse in orderto answer my research question “What could those specialised teams improve?” The first partof this work is theoretical and provides information on the structure of the Belgian and SwedishPolice. This will be followed by a brief discussion of the role of Europol and Interpol in thefight against child sexual abuse, as well as a short chapter on cryptography to understand howinternational communications are conducted.The second part is about the practical part of this work. To collect the necessary information,the Belgian Police opened their doors, specifically to the Child Abuse section, for three days.They introduced their teams, equipment, current and closed investigations, and answered a pre-prepared list of questions.It was not possible to meet with the Swedish Police in person. Therefore, it was decided togather the necessary information through interviews. The questions were based on the timespent with the Belgian team.Once the information was gathered, it was compared, and a score was assigned to each responseprovided by Belgium and Sweden to give an idea of their accuracy and completeness.Subsequently, a list of recommendations was formulated with the aim of proposing areas forimprovement.The statistics presented at the end of this thesis allow for a visualization of the evolution of thenumber of assaults and investigations opened and resolved.

Supervised learning algorithms have been used for automatic text categoriza- tion with very good results. But supervised learning requires a large amount of manually labeled training data and this is a serious limitation for many practical applications. Keyword-based text categorization does not require manually la- beled training data and has therefore been presented as an attractive alternative to supervised learning. The aim of this study is to explore if there are other li- mitations for using keyword-based text categorization in industrial applications. This study also tests if a new lexical resource, based on the paradigmatic rela- tions between words, could be used to improve existing keyword-based text ca- tegorization algorithms. An industry motivated use case was created to measure practical applicability. The results showed that none of five examined algorithms was able to meet the requirements in the industrial motivated use case. But it was possible to modify one algorithm proposed by Liebeskind et.al. (2015) to meet the requirements. The new lexical resource produced relevant keywords for text categorization but there was still a large variance in the algorithm’s capaci- ty to correctly categorize different text categories. The categorization capacity was also generally too low to meet the requirements in many practical applica- tions. Further studies are needed to explore how the algorithm’s categorization capacity could be improved.

Computer vision for autonomous vehicles is a widely studied and quickly evolving field of research. It can be difficult to know when one technology is performs better or worse, and in which scenario it's worth it to spend the additional resources on LiDAR. This paper defines criteria for comparison between the two technologies, compares the criteria in different environments designed to test each of the technologies' weaknesses and finally evaluates the performances. After evaluating the performances of the defined criteria the paper concludes that despite the smaller price tag of the camera, they can be a very viable choice overall with an above average performance in the selected environments, as well as maintaining some benefits that LiDAR can't handle. However overall, if the budget allows, LiDAR performs generally better in all environments.

No description available.

The main focus of this thesis is to research the usability and reliability of the patient-focused journal system Cosmic that is created by the company Cambio. The system is widely used in Sweden, increasing the pressure on the system to perform well and meet the needs of the healthcare personnel. This thesis identifies potential areas of improvement in the electronic journal system. The study is based on a literature and an empirical study, which is based on a thematic and content analysis of focus group interview results and a User Experience (UX) case study. The study found that participants reported numerous usability flaws, such as extensive clicking, imprecise warnings, and a design that was difficult to understand, making finding vital parameters difficult. It was also discovered that there was a need for more customized and personalized solutions, particularly in the patient overview. Poor response times and frequent forced log out were also mentioned as reliability issues. Through the data collected from the interviews and literature study, a new design of the patient view is made and presented via a survey to the users, in which the users give their ratings on the new design. The result from this survey suggests that features that improve usability, customization, and accessibility are valued by the users, which may seem obvious. However, it is crucial to acknowledge that the current system lacks these essential elements. The thesis underlines the importance of reliability in making sure healthcare workers have immediate and unrestricted access to patient information. These concerns with usability and reliability highlight the need for an updated design that increases the effectiveness and efficiency of healthcare systems. These issues have been identified in previous research and still persist, underlining their severity and the need for action to address them.

Sökproblem förekommer i många olika områden. I denna rapport har vi tacklat problemet med atthitta en person som gått vilse. Vi har utvecklat en sökalgoritm vars uppgift är att hitta personengenom att använda erfarenheter från tidigare sökuppdrag. Sökalgoritmen baserar sig på A*sökalgoritmen och sannolikheter. A* sökalgoritmen använder sig av heuristik som vägledersökningen och heuristiken bygger på erfarenhet. Vi har implementerat ett antal sökalgoritmer, treolika A* och två oinformerade sökalgoritmer för att utvärdera och jämföra deras söktid. Vi haräven utvärderat hur snabbt personen hittas med flera sökare och hur användningen av ledtråd somhittas under sökningen förbättrar sökalgoritmen. Vi använder en simulator som förser oss medvärdena som behövs för A* versionerna, vi kör ett stort antal tester med olika sökningsområdenoch mäter hur lång tid det tar att hitta personen för att skapa ett genomsnittligt resultat för de olikasökalgoritmerna. De två oinformerade sökalgoritmerna som vi implementerade söker på breddenrespektive djupet först. De har använts som benchmarking för att jämföra med de olika A*algoritmerna och analysera resultatet. Alla A* versioner var snabbare med att hitta barnet jämförtmed bredden och djupet först sökningsalgoritmerna. Användningen av fler sökare gjorde att tidensom behövdes för ett vanligt sökuppdrag med de övriga A* sökalgoritmerna reducerades avsevärt.I vårt fall med 3 sökare reducerades tiden till en tredjedel. Användningen av en ledtråd förbättradede två första A* algoritmerna med mer än 5%.

As a student project at Halmstad University, CatFish aims to make water quality measurement easier by the development of a set of cooperating measurement drones in and under water and up in the air. To minimize human effort during the operation of the CatFish system, the drones shall act autonomously to reach given sets of target coordinates on their own to fulfill desired measurement tasks on arrival. This thesis presents the development of an autonomy system for the water surface drone from development, integration and testing in a prototype in different configurations and under continuous improvement with the goal to identify and implement an ideal system solution. In addition to this basic autonomy system, the second major part of the presented work consists of the development of an object detection and tracking system for increased safety of the floating drone and its environment. Confronted with very limited computational power which is a property of most embedded systems, this thesis puts a strong focus on the investigation of state-of-the-art technology in the area of computer vision to identify the most promising approaches in the constraint environment. To achieve the goal of creating a useful solution, the development of different object detection and object tracking algorithms and combinations of both methods and their evaluation was done with a concentration on algorithm speed rather then accuracy in the first place. The outcome is a first set of possible implementations of object detection and tracking models, accompanied with recommendations for further speed improvements, to create an obstacle avoidance system that can extend the developed autonomy system of the water surface drone of the CatFish project in the future.

Volumetric Lighting is a collective term for visual phenomena that occur due to how light interacts inside so-called participating media, and it accounts for many recognizable effects such as fog or light shafts. Because it isvery computationally expensive, it has been an importantproblem within computer graphics to calculate volumetriclighting, both accurately and efficiently. Ray Marching is a technique that has been used extensively in non real-time applications to compute volumetriclighting and has recently been adapted for real time applications by use of the GPU. In this thesis we implement andevaluatevolumetric ray marching with ordered dithering. The results show how ordered dithering yields significant performance improvements, retaining high quality while lowering the number of samples. We conclude that with ordered dithering, volumetric ray marching is a suitable approach for real time volumetric ray marching on the GPU and we discuss both important additional optimizationsand how ordered dithering will likely remain important in future ray marching implementations. / Volumetriskt ljus är en term som beskriver visuella fenomen som uppstår från hur ljus interagerar inuti material som kan bära ljus. Hur ljuset absorberas eller ändrar riktning då det färdas igenom material ger upphov till många bekanta fenomen såsom dimma, moln eller eld. Eftersom volumetriskt ljus är dyrt att beräkna så har det varit ett viktigt problem inom datorgrafik hur man effektivt simulerar denna typ av ljustransport. Ray Marching är en metod som har använts mycket inom bland annat filmindustrin där man inte har en hård gräns på beräkningstiden, men metoden har med hjälp av grafikkortets parallelliseringsförmåga också börjat appliceras för realtidsapplikationer såsom datorspel. I denna rapport så utforskar vi en optimeringsmetod till grafikkortsbaserad ray marching som kallas för ordered dithering. Resultaten visar hur optimeringsmetoden ger stor prestandaförbättring genom att placera samplingspunkter mer effektivt, utan signifikant försämring av kvalité. Resultaten styrker hur den valda algoritmen är en lämplig algoritm för att åstadkomma volumetriskt ljus i realtid. Vi diskuterar också hur optimeringsmedoten troligtvis även i framtiden kommer spela en viktig roll i att nå acceptabel prestandainom grafikkortsbaserad ray marching.

Using presentation software such as PowerPoint or Keynote to sup- port lectures and presentations has become ubiquitous in both academia and industry. However, designing the visuals of presentation slides is a time-consuming and laborious task; repetitive steps are required for selecting templates, organizing objects, and optimizing layouts. To al- leviate these laborious works and to allow users to focus on preparing the contents of presentations, we present SmartPPT, a framework that supports the automatic generation of presentation slides from textual outline. We built a Recommender System model inside the framework that could pick up slide templates for input textual outline. To test its functionality and efficiency, two sets of user study procedures were conducted and shown that SmartPPT is time efficient in generating slides and it outperforms in user satisfaction when compared to user- handcrafted slides and the baseline condition, which was PowerPoint suggested templates. / Att använda presentationsprogram som PowerPoint eller Keynote för att stödja föreläsningar och presentationer har blivit allestädes närvarande både i akademin och industrin. Att utforma visuella bilder av presentationsmaterial är emellertid en tidskrävande och mödosam uppgift; upprepade steg krävs för att välja mallar, organisera objekt och optimera layouter. För att underlätta dessa mödosamma arbeten och låta användarna fokusera på att förbereda innehållet i presentationer presenterar vi SmartPPT, ett ramverk som stöder den automatiska generationen av presentationsbilder från en textvy. Vi byggde en Recommender System-modell inuti ramen som kunde plocka upp bildmallar för inmatning av textrutor. För att testa dess funktionalitet och effektivitet genomfördes två uppsättningar av användarstudieprocedurer och visade att SmartPPT är tidseffektivt för att generera bilder och användarna blev nöjdare jämfört med när man gjort handgjorda bilder eller använt PowerPoint-föreslagna mallar.