A cada dia aumenta o número de vídeos disponíveis no mundo. Por exemplo, há uma vasta quantidade de sites de vídeos disponíveis na Web e serviços de Vídeo Sob Demanda além de dispositivos que fazem a gravação de vídeos automaticamente, conhecidos como Personal Video Recorders, 24 horas por dia. Isso pode ocasionar um problema ao usuário: a sobrecarga de conteúdo em formato de vídeo. Uma das maneiras de se tratar tal problema consiste no uso de sistemas de recomendação, os quais filtram o conteúdo com o objetivo de entregar o que for mais interessante ao usuário. A abordagem típica utilizada pelos sistemas atuais consiste em um sistema de recomendação híbrido, i.e., que utiliza tanto filtragem baseada em conteúdo quanto filtragem colaborativa, minimizando os problemas que tais abordagens possuem individualmente. Adicionalmente, com o objetivo de melhorar a recomendação ou de criar novas formas de recomendação, têm sido apresentadas novas abordagens, tais como sistemas de recomendação utilizando dados de redes sociais, computação afetiva, tags, entre outros. Este trabalho tem como objetivo apresentar uma abordagem inovadora, a qual utiliza pontos de interesse em vídeo de usuários (ou seja, os segmentos dos vídeos que eles mais gostam ou que mais se interessam) para melhorar a acurácia de predição de sistemas de recomendação de vídeo que utilizam filtragem colaborativa baseados na abordagem usuário-usuário. Na abordagem proposta, os usuários participam de forma mais ativa e mais interativa ao marcarem seus pontos de interesse. Para avaliação de tal abordagem proposta foi realizada uma avaliação experimental em termos de acurácia de predição de avaliação; pela qual constatou-se que houve melhora na predição de avaliação do sistema de recomendação. Tal melhora está diretamente relacionada com o nível de participação das pessoas na marcação de pontos de interesse. / Every day the number of videos available in the world increases. For example, there is a vast amount of video sites available on the Web, Video On Demand services, as well as devices that records videos automatically, known as Personal Video Recorders, 24 hours a day. It may create a problem for the user: the overload of content in video format. One of the ways to treat such problem is the use of recommender systems, which filter the content in order to deliver what is most interesting to the user. The typical approach is to present a hybrid recommender system, i.e., that uses both contentbased filtering and collaborative filtering, minimizing the problems that these approaches have individually. Additionally, in order to improve the recommendation or to create new approaches of recommendation, has been given new approaches such as systems using data from social networks, affective computing, tags, etc. This paper aims to present an innovative approach, which uses points of interest (POI) in video of users (i.e., video segments best liked or most interested by them) to augment the prediction accuracy of video recommender systems with collaborative filtering based in the useruser approach. In the proposed approach, users participate more actively and more interactively to mark their points of interest. To evaluate this proposed approach an experimental evaluation was performed in terms of accuracy of ratings predictions; in which it was verified that there was an improvement in ratings prediction accuracy of the recommendation system. This improvement is directly related to the level of participation of people in marking points of interest.

A cada dia aumenta o número de vídeos disponíveis no mundo. Por exemplo, há uma vasta quantidade de sites de vídeos disponíveis na Web e serviços de Vídeo Sob Demanda além de dispositivos que fazem a gravação de vídeos automaticamente, conhecidos como Personal Video Recorders, 24 horas por dia. Isso pode ocasionar um problema ao usuário: a sobrecarga de conteúdo em formato de vídeo. Uma das maneiras de se tratar tal problema consiste no uso de sistemas de recomendação, os quais filtram o conteúdo com o objetivo de entregar o que for mais interessante ao usuário. A abordagem típica utilizada pelos sistemas atuais consiste em um sistema de recomendação híbrido, i.e., que utiliza tanto filtragem baseada em conteúdo quanto filtragem colaborativa, minimizando os problemas que tais abordagens possuem individualmente. Adicionalmente, com o objetivo de melhorar a recomendação ou de criar novas formas de recomendação, têm sido apresentadas novas abordagens, tais como sistemas de recomendação utilizando dados de redes sociais, computação afetiva, tags, entre outros. Este trabalho tem como objetivo apresentar uma abordagem inovadora, a qual utiliza pontos de interesse em vídeo de usuários (ou seja, os segmentos dos vídeos que eles mais gostam ou que mais se interessam) para melhorar a acurácia de predição de sistemas de recomendação de vídeo que utilizam filtragem colaborativa baseados na abordagem usuário-usuário. Na abordagem proposta, os usuários participam de forma mais ativa e mais interativa ao marcarem seus pontos de interesse. Para avaliação de tal abordagem proposta foi realizada uma avaliação experimental em termos de acurácia de predição de avaliação; pela qual constatou-se que houve melhora na predição de avaliação do sistema de recomendação. Tal melhora está diretamente relacionada com o nível de participação das pessoas na marcação de pontos de interesse. / Every day the number of videos available in the world increases. For example, there is a vast amount of video sites available on the Web, Video On Demand services, as well as devices that records videos automatically, known as Personal Video Recorders, 24 hours a day. It may create a problem for the user: the overload of content in video format. One of the ways to treat such problem is the use of recommender systems, which filter the content in order to deliver what is most interesting to the user. The typical approach is to present a hybrid recommender system, i.e., that uses both contentbased filtering and collaborative filtering, minimizing the problems that these approaches have individually. Additionally, in order to improve the recommendation or to create new approaches of recommendation, has been given new approaches such as systems using data from social networks, affective computing, tags, etc. This paper aims to present an innovative approach, which uses points of interest (POI) in video of users (i.e., video segments best liked or most interested by them) to augment the prediction accuracy of video recommender systems with collaborative filtering based in the useruser approach. In the proposed approach, users participate more actively and more interactively to mark their points of interest. To evaluate this proposed approach an experimental evaluation was performed in terms of accuracy of ratings predictions; in which it was verified that there was an improvement in ratings prediction accuracy of the recommendation system. This improvement is directly related to the level of participation of people in marking points of interest.

A cada dia aumenta o número de vídeos disponíveis no mundo. Por exemplo, há uma vasta quantidade de sites de vídeos disponíveis na Web e serviços de Vídeo Sob Demanda além de dispositivos que fazem a gravação de vídeos automaticamente, conhecidos como Personal Video Recorders, 24 horas por dia. Isso pode ocasionar um problema ao usuário: a sobrecarga de conteúdo em formato de vídeo. Uma das maneiras de se tratar tal problema consiste no uso de sistemas de recomendação, os quais filtram o conteúdo com o objetivo de entregar o que for mais interessante ao usuário. A abordagem típica utilizada pelos sistemas atuais consiste em um sistema de recomendação híbrido, i.e., que utiliza tanto filtragem baseada em conteúdo quanto filtragem colaborativa, minimizando os problemas que tais abordagens possuem individualmente. Adicionalmente, com o objetivo de melhorar a recomendação ou de criar novas formas de recomendação, têm sido apresentadas novas abordagens, tais como sistemas de recomendação utilizando dados de redes sociais, computação afetiva, tags, entre outros. Este trabalho tem como objetivo apresentar uma abordagem inovadora, a qual utiliza pontos de interesse em vídeo de usuários (ou seja, os segmentos dos vídeos que eles mais gostam ou que mais se interessam) para melhorar a acurácia de predição de sistemas de recomendação de vídeo que utilizam filtragem colaborativa baseados na abordagem usuário-usuário. Na abordagem proposta, os usuários participam de forma mais ativa e mais interativa ao marcarem seus pontos de interesse. Para avaliação de tal abordagem proposta foi realizada uma avaliação experimental em termos de acurácia de predição de avaliação; pela qual constatou-se que houve melhora na predição de avaliação do sistema de recomendação. Tal melhora está diretamente relacionada com o nível de participação das pessoas na marcação de pontos de interesse. / Every day the number of videos available in the world increases. For example, there is a vast amount of video sites available on the Web, Video On Demand services, as well as devices that records videos automatically, known as Personal Video Recorders, 24 hours a day. It may create a problem for the user: the overload of content in video format. One of the ways to treat such problem is the use of recommender systems, which filter the content in order to deliver what is most interesting to the user. The typical approach is to present a hybrid recommender system, i.e., that uses both contentbased filtering and collaborative filtering, minimizing the problems that these approaches have individually. Additionally, in order to improve the recommendation or to create new approaches of recommendation, has been given new approaches such as systems using data from social networks, affective computing, tags, etc. This paper aims to present an innovative approach, which uses points of interest (POI) in video of users (i.e., video segments best liked or most interested by them) to augment the prediction accuracy of video recommender systems with collaborative filtering based in the useruser approach. In the proposed approach, users participate more actively and more interactively to mark their points of interest. To evaluate this proposed approach an experimental evaluation was performed in terms of accuracy of ratings predictions; in which it was verified that there was an improvement in ratings prediction accuracy of the recommendation system. This improvement is directly related to the level of participation of people in marking points of interest.

Στην εργασία αυτή ασχολούμαστε με την αξιοποίηση των δεδομένων από διαφορετικά κοινωνικά δίκτυα (πιο συγκεκριμένα από Foursquare και Facebook) με σκοπό να ταυτοποιήσουμε τις ίδιες τοποθεσίες (ή αλλιώς σημεία ενδιαφέροντος) που έχουν εισαχθεί σε αυτά τα δίκτυα. Το πρόβλημα της ταυτοποίησης είναι σημαντικό να λυθεί διότι έτσι θα μπορούσε να αποκτηθεί μια καλύτερη εικόνα για τις αλληλεπιδράσεις των χρηστών με το φυσικό περιβάλλον με τη χρήση των μέσων κοινωνικής δικτύωσης (social data). Αυτό σημαίνει ταυτόχρονα και καλύτερη ανάλυση και αξιοποίηση αυτών δεδομένων, αφού θα έχουμε αναγνωρίσει μεγάλο μέρος των κοινών σημείων ενδιαφέροντος από ετερογενή σύνολα δεδομένων από τα μέσα κοινωνικής δικτύωσης. Μια λύση στο πρόβλημα είναι η χρήση των αλγορίθμων μηχανικής μάθησης, που θα αποφασίζουν αν ένα ζεύγος σημείων αντιπροσωπεύει το ίδιο σημείο ενδιαφέροντος. / In this paper we deal with the exploitation of data from different social networks (more specifically from Foursquare and Facebook) in order to identify the same locations (or landmarks ) introduced in these networks . The problem of identification is important to solve it so he could get a better picture of the user interactions with the natural environment through the use of social media (social data). This means simultaneously and better analysis and use of such data , since we recognize much of the common points of interest from heterogeneous datasets from social media . One solution to this problem is the use of machine learning algorithms , which will decide whether a pair of points represents the same point of interest .

Grâce aux technologies web et mobiles, le partage de données entre utilisateurs a considérablement augmenté au cours des dernières années. Par exemple, les utilisateurs peuvent facilement enregistrer leurs trajectoires durant leurs déplacements quotidiens avec l'utilisation de récepteurs GPS et les mettre en relation avec les trajectoires d'autres utilisateurs. L'analyse des trajectoires des utilisateurs au fil du temps peut révéler des habitudes et préférences. Cette information peut être utilisée pour recommander des contenus à des utilisateurs individuels ou à des groupes d'utilisateurs avec des trajectoires ou préférences similaires. En revanche, l'enregistrement de points GPS génère de grandes quantités de données. Par conséquent, les algorithmes de clustering sont nécessaires pour analyser efficacement ces données. Dans cette thèse, nous nous concentrons sur l'étude des différentes solutions pour analyser les trajectoires, extraire les préférences et identifier les intérêts similaires entre les utilisateurs. Nous proposons un algorithme de clustering de trajectoires GPS. En outre, nous proposons un algorithme de corrélation basée sur les trajectoires des points proches entre deux ou plusieurs utilisateurs. Les résultats finaux ouvrent des perspectives intéressantes pour explorer les applications des réseaux sociaux basés sur la localisation. / Sharing of user data has substantially increased over the past few years facilitated by sophisticated Web and mobile applications, including social networks. For instance, users can easily register their trajectories over time based on their daily trips captured with GPS receivers as well as share and relate them with trajectories of other users. Analyzing user trajectories over time can reveal habits and preferences. This information can be used to recommend content to single users or to group users together based on similar trajectories and/or preferences. Recording GPS tracks generates very large amounts of data. Therefore clustering algorithms are required to efficiently analyze such data. In this thesis, we focus on investigating ways of efficiently analyzing user trajectories, extracting user preferences from them and identifying similar interests between users. We demonstrate an algorithm for clustering user GPS trajectories. In addition, we propose an algorithm to correlate trajectories based on near points between two or more users. The final results provided interesting avenues for exploring Location-based Social Network (LBSN) applications.

Dans cette thèse, une nouvelle représentation des images couleur basée sur une généralisation du signal analytique est introduite. En utilisant l'analogie entre les conditions de Cauchy-Riemann, qui définissent le caractère holomorphe d'une fonction, et l'équation de Dirac dans l'algèbre de Clifford R_{5,0}, un système d'équations dont la solution est le signal monogène couleur est obtenu. Ce signal est notamment basé sur des noyaux de Riesz ainsi que de Poisson 2D, et une représentation polaire, basée sur un produit géométrique, peut lui être associée. Les applications envisagées reposent majoritairement sur cette représentation polaire et sur les informations de couleur et de structures locales s'y rattachant. Des problématiques liées au flot optique couleur, à la segmentation couleur multi-échelle, au suivi d'objets couleur et à la détection de points d'intérêt sont abordées. En ce qui concerne le flot optique, nous nous intéressons à l'extraction du mouvement d'objets d'une certaine couleur en remplaçant la contrainte de conservation de l'intensité par une contrainte de conservation d'angles. Pour la segmentation, une méthode de détection de contours basée sur de la géométrie différentielle et plus particulièrement sur la première forme fondamentale d'une surface, est proposée afin de déterminer les contours d'objets d'une couleur choisie. Pour le suivi d'objets, nous définissons un nouveau critère de similarité utilisant le produit géométrique que nous insérons dans un filtrage particulaire. Enfin, nous resituons la définition du détecteur de Harris dans le cadre de la géométrie différentielle en faisant le lien entre ce dernier et une version "relaxée" du discriminant du polynôme caractéristique de la première forme fondamentale. Ensuite nous proposons une nouvelle version multi-échelle de ce détecteur en traitant le paramètre d'échelle comme une variable d'une variété de dimension 3. / In this thesis, a novel framework for color image processing is introduced based on the generalization of the analytic signal. Using the analogy between the Cauchy-Riemann conditions and the Dirac equation in the Clifford algebra R_{5,0}, a system of equations which leads to the color monogenic signal is obtained. This latter is based on the Riesz and 2D Poisson kernels, and a polar representation based on the geometric product can be associated to this signal. Some applications using color and local structure information provided by the polar representation are presented. Namely, color optical flow, color segmentation, color object tracking and points of interest are developed. Extraction of optical flow in a chosen color is obtained by replacing the brightness constancy assumption by an angle constancy. Edge detection is based on the first fundamental form from differential geometry in order to segment object in a predefined color. Object tracking application uses a new similarity criterion defined by geometric product of block of vectors. This latter is viewed as the likelyhood measure of a particle filter. Last part of the thesis is devoted to the definition of the Harris detector in the framework of differential geometry and a link between this definition and a relaxed version of the characteristic polynomial discriminant of the first fundamental form is given. In this context, a new scale-space detector is provided as the result of handling the scale parameter as a variable in a 3-manifold.

The Witcher 3: Wild Hunt is one of the best selling open world video games in recent years. Thanks to the game’s popularity, other game designers might want to replicate its world design methods. In an interview from 2017, the game’s developers have explained that they have used a system called “the rule of 40 seconds” to ensure that the player stumbles upon something interesting every 40 seconds of exploration. In this research, the authors conducted a test to check how accurately the rule was implemented; they did this by analysing footage of gameplay from four different YouTube content creators, on two of the game’s maps. The research found that a Rule does exist, however, it is greatly influenced by playstyles and only accurate for certain players. It concludes that designers looking to implement the rule, can reliably do so, but they should take playing styles and player movement speed into consideration when testing their rule. / The Witcher 3: Wild Hunt är ett av de bäst säljande öppna världsspel de senaste åren. Tack vare spelets popularitet kanske andra speldesigners vill kopiera dess värld designmetoder. I en intervju från 2017 har spelets utvecklare förklarat att de har använt ett system som kallas ”regeln om 40 sekunder” för att säkerställa att spelaren snubblar på något intressant var fjärde sekund av utforskningen. I denna forskning genomförde författarna ett test för att kontrollera hur korrekt regeln implementerades; de gjorde det genom att analysera bilder från fyra olika innehållsskapare på YouTube på två av spelets kartor. Forskningen visade att det finns en regel, men den påverkas av spelstilar och är endast korrekt för vissa spelare. Det drar slutsatsen att designers som vill implementera regeln kan göra det på ett tillförlitligt sätt, men de bör ta hänsyn till spelstilar och spelarens rörelsehastighet när de testar deras regel.

This document deals with experimental verifying to use machine learning algorithms AdaBoost or WaldBoost to make classifier, that is able to find point in the second picture that matches original point in the first picture. This work also depicts finding points of interest in image as a first step of finding correspondence. Next there are described some descriptors of points of interest. Corresponding points could be useful for 3D modeling of shooted scene.

The open world game The Witcher 3: Wild Hunt (2015) has been proven to follow a 40 second rule when placing points of interest for players to explore. The developers of the game said in an interview that the rule existed, and this was proven by a study made in 2021 by Cojanu and Jaber (2021). The 40 second rule means that whichever direction the player goes in, they will encounter a point of interest within 40 seconds. This study asks the question if Genshin Impact (2020) is following the 40 second rule. The researchers analyzed footage from four YouTube content creators to see if the rule was implemented in the game. It was found that Genshin Impact (2020) follows the 40 second rule since the points of interest were about 10 seconds apart. How the player decides to play the game slightly impacts the frequency of the points of interest and could be taken into account when testing the 40 seconds rule. It was found that points of interest found were more random at the start of the game and more intentionally found by the player later in the game. / Öppen världs spelet The Witcher 3: Wild Hunt (2015) har visat sig följa en 40 sekunders regel när det gäller placeringen av intressanta platser för spelare att utforska. Spelets utvecklare sa i en intervju att regeln fanns, och detta bevisades genom en studie utförd 2021 av Cojanu och Jaber (2021). Regeln innebär att oavsett vilken riktning spelaren går i så kommer de att stöta på en intressant plats inom 40 sekunder. Denna studie undersöker om Genshin Impact (2020) följer 40 sekunders regeln. Forskarna analyserade videomaterial från fyra YouTube innehållsskapare för att se om regeln implementerades i spelet. Det visades att Genshin Impact (2020) följer 40 sekunders regeln, eftersom de intressanta platserna låg ungefär 10 sekunder ifrån varandra. Hur spelaren väljer att spela spelet påverkar lite grann frekvensen av de intressanta platser och detta kan tas i beaktning när man testar 40 sekunders regeln. Det visade sig att de intressanta platserna var mer slumpmässiga i början av spelet och mer avsiktligt hittade av spelaren senare i spelet.

Open-world videogames inherently allow players a significant amount of freedomwhen traversing the in-game world. Due to this fact, level designers for such games are taskedwith the challenge of maintaining engagement and motivating exploration without drasticallyinfringing on their player agency. With this study, the authors analyzed three popular open-world games via a comparative formal analysis to determine the presence and validity of thePOI Diversity Rule, a conceptual rule for effective open-world level design proposed bySkobelev (2023), which recommends having at least three points of interests on the player’sline of horizon, offering different gameplay experiences. The authors performed this study byplotting out a route known as a “critical path”, taking panoramic screenshots within 30 secondintervals and analyzing them. The results showed that the rule is largely prevalent in all threegames. However, the conditions of fulfillment differ in terms of what categories are mostcommon, and whether static or dynamic POI are most prevalent. / Open world-datorspel ger spelare en hög grad av frihet att resa runt i spelvärlden. Pågrund av detta står leveldesigners för sådana spel inför utmaningen att upprätthållaengagemang och motivera utforskning inom spelvärlden utan att allvarligt begränsa spelarnashandlingsfrihet. I denna studie analyserades tre populära open world-spel via en jämförandeformell analys för att avgöra närvaron och giltigheten av POI Diversity Rule, en konceptuellregel för effektiv open world-leveldesign föreslagen av Skobelev (2023). Regelnrekommenderar att det bör finnas minst tre intressepunkter inom spelarens synligahorisontlinje, som erbjuder olika spelupplevelser. Författarna genomförde denna studiegenom att plotta in en rutt kallad "critical path", ta panoramabilder med 30 sekundersintervaller och analysera dem. Resultaten visade att regeln är till stor del närvarande i alla trespel. Men de specifika villkoren gällande dess uppfyllelse skiljer sig åt när det kommer tillvilka kategorier som är vanligast, och huruvida statiska eller dynamiska intressepunkter ärmest framträdande. / <p></p><p></p>