FER Repository
Not a member yet
12128 research outputs found
Sort by
Plan upravljanja istraživačkim podacima projekta DESCARTES - Određivanje karakteristika učenika korištenjem edukacijskih robotskih tehnologija
No full textPlan upravljanja istraživačkim podacima projekta DESCARTES - Određivanje karakteristika učenika korištenjem edukacijskih robotskih tehnologija, HRZZ-IP-2022-10-1915
PLAN UPRAVLJANJA ISTRAŽIVAČKIM PODACIMA - Razvoj hiperheuristika za relokaciju kontejnera
No full textA Framework for Improving Accessibility of Serious Games in Handheld Augmented Reality Based on User Interaction Data
No full textSerious games combined with augmented reality (AR) can positively impact user’s motivation and learning experience. To ensure that users with different needs can use them effectively, we must ensure the accessibility of such solutions. The goal of this paper is to show that game-learning analytics can be utilized to identify issues for people with various needs and define ways to improve the accessibility of AR-based serious games. This paper presents a framework based on analysing automatically collected user interaction data in handheld AR-based serious games to adapt them to the user in the context of accessibility. An experiment was conducted with people with disabilities and people of different ages using the established framework. The focus of the experiment results analysis in this paper is on the evaluation of interaction mechanisms adapted to users based on their interaction data and preferences. The results showed that for most participants the assigned interaction mechanisms for virtual object transformation tasks (translation, scaling, rotation) in AR were satisfactory. The framework proposed in this paper provides the basis for further research in this field concerning the use of advanced analytics that enable intelligent adaptation to the individual user
Digital catalog of products
No full textOvaj rad opisuje izradu i razvoj web aplikacije koja funkcionira kao digitalni katalog proizvoda. Kroz rad su detaljno razmotreni funkcionalni i nefunkcionalni zahtjevi sustava, čime se definiraju ključne operacije i kvalitativni aspekti aplikacije. U radu su opisane i implementirane glavne komponente, a posebna pažnja posvećena je integraciji različitih tehnologija i alata. Za razvoj aplikacije korištene su moderne tehnologije od kojih je React.js odabran za razvoj interaktivnog korisničkog sučelja, dok je poslužiteljev sloj izgrađen na .NET platformi s Entity Frameworkom, što omogućava učinkovito upravljanje poslovnom logikom i bazom podataka. Za pohranu podataka koristi se PostgreSQL. Dodatno su u radu opisane i tablice izgrađene baze podataka i njen ERD dijagram. Rad sadrži primjere koda i slike kroz koje se opisuje napravljena aplikacija i komunikacija između klijentskog i poslužiteljevog sloja.This paper describes the design and development of a web application that functions as a digital product catalog. The study examines in detail the system's functional and non-functional requirements, thereby defining the key operations and qualitative aspects of the application. The paper outlines and implements the main components, with particular emphasis on the integration of various technologies and tools. Modern technologies were used for the development of the application, with React.js chosen for creating the interactive user interface and the server side built on the .NET platform with Entity Framework, enabling effective management of business logic and the database. PostgreSQL is used for data storage. Additionally, the paper describes the tables of the constructed database along with its ERD diagram. The work includes code examples and screenshots that illustrate the developed application and the communication between the client and server layers
A user interface for creating and storing configuration data for a WireGuard virtual private network
No full textTema ovog rada je istraživanje i razvijanje korisničkog sučelja za kreiranje i pohranu
konfiguracijskih podataka za program WireGuard, koji omogućava sigurno stvaranje
virtualnih privatnih mreža (VPN) korištenjem napredne kriptografije. Provedena implementacija
omogućava autoriziranim korisnicima samostalno dodavanje i konfiguriranje novih uređaja,
koristeći standardne autentifikacijske mehanizme kao što su AAI@EduHr i Google Single
Sign On. Posebna pažnja posvećena je sigurnosnoj zaštiti sustava. U sklopu rada, simulirani
su standardni napadi kako bi se provjerila otpornost i sigurnost sustava. Rezultati testiranja
pokazali su da sustav uspješno odolijeva većini poznatih sigurnosnih prijetnji, što potvrđuje
njegovu pouzdanost i sigurnost u stvarnim uvjetima.The topic of this paper is the research and development of a user interface for creating
and storing configuration data for the WireGuard program, which enables secure creation
of virtual private networks (VPN) using advanced cryptography. The implemented solution
allows authorized users to independently add and configure new devices using standard
authentication mechanisms such as AAI@EduHr and Google Single Sign-On. Special
attention has been given to the security protection of the system. Standard attacks were
simulated as part of the study to verify the system’s resilience and security. The test
results showed that the system successfully withstands most known security threats,
confirming its reliability and security in real-world conditions
Object Detection on RISC-V Platform
No full textRačunalni vid i detekcija objekata postali su ključni segmenti modernih sustava umjetne inteligencije, pri čemu se sve više istražuju alternativne hardverske platforme koje omogućuju njihovu izvedbu. Ovaj rad analizira mogućnosti primjene RISC-V arhitekture u kontekstu računalnog vida, s posebnim naglaskom na izvedbu detekcije objekata na VisionFive 2 jednopločnom računalu.
U prvom dijelu rada predstavljene su temeljne značajke RISC-V arhitekture te usporedba s popularnijim ARM i x86 rješenjima. Također se istražuju tehničke specifikacije VisionFive 2 uređaja, njegova kompatibilnost s modernim softverskim bibliotekama te ograničenja koja proizlaze iz nedovršene GPU podrške.
Eksperimentalni dio rada obuhvaća implementaciju i testiranje YOLO i MobileNet SSD modela na VisionFive 2. Analizirane su performanse u pogledu brzine inferencije na CPU-u te potencijalni dobitci uz buduću GPU akceleraciju. Rezultati pokazuju da iako VisionFive 2 može izvršavati osnovne zadatke računalnog vida, postoje značajna ograničenja u odnosu na konkurentske SBC uređaje poput Raspberry Pi-a, prvenstveno zbog nedostatka optimiziranih drivera i podrške za hardversku akceleraciju.
Zaključni dio rada ističe ključne izazove i moguća poboljšanja za buduću implementaciju računalnog vida na RISC-V platformama. Iako se RISC-V pokazuje kao perspektivna alternativa zatvorenim arhitekturama, potrebno je daljnje usavršavanje softverskog ekosustava kako bi se postigla bolja integracija s modernim AI modelima.Computer vision and object detection have become essential components of modern artificial intelligence systems, prompting increased exploration of alternative hardware platforms for their implementation. This paper examines the feasibility of utilizing RISC-V architecture for computer vision applications, with a particular focus on object detection performance on the VisionFive 2 single-board computer.
The first part of the paper introduces the fundamental features of RISC-V architecture and compares it with more widely adopted ARM and x86 solutions. Additionally, the technical specifications of the VisionFive 2 device are analyzed, including its compatibility with modern software libraries and the limitations arising from incomplete GPU support.
The experimental section involves the implementation and benchmarking of YOLO and MobileNet SSD models on VisionFive 2. The study evaluates inference performance using the CPU and explores the potential benefits of future GPU acceleration. The results indicate that while VisionFive 2 can handle basic computer vision tasks, it faces significant limitations compared to competing SBC devices like the Raspberry Pi, primarily due to the lack of optimized drivers and hardware acceleration support.
The conclusion highlights key challenges and potential improvements for the future development of computer vision applications on RISC-V platforms. Although RISC-V presents a promising alternative to proprietary architectures, further enhancements in its software ecosystem are required to achieve better integration with modern AI models
Monitoring Student Engagement Using Smart Watches
No full textU ovom radu opisana je primjena pametnih satova na nastavi. Prikupljanje podataka
odvijalo se u tri faze u različitim uvjetima: na nastavi, tijekom robotičke radionice gdje
su promatrači pratili događaje te tijekom ispitnog tjedna. Za vrijeme prikupljanja
podataka svaki je student nosio pametni sat označen brojem. Podaci snimljeni s pomoću
pametnih satova obrađivani su i analizirani u programskom jeziku Python, a rezultati su
vizualizirani u obliku grafova i tablica. Na kraju rada prezentirani su i komentirani
rezultati i zaključci.This paper describes the use of smartwatches in education. Data collection took place in
three phases in different conditions: in class, during a robotic workshop where observers
followed the events, and during an exam week. During data collection, each student wore
a smartwatch marked with a number. The data recorded using smartwatches were
processed and analyzed in the Python programming language, and the results were
visualized in the form of graphs and tables. At the end of the thesis, the results and
conclusions were presented and commented on
Platform for predicting the production of solar power plants in Croatia
No full textOvaj diplomski rad bavi se razvojem platforme za predviđanje proizvodnje solarnih
elektrana u Hrvatskoj s ciljem poboljšanja učinkovitosti i optimizacije upravljanja
solarnim elektranama. Platforma je razvijena koristeći moderne web tehnologije, a
fokus rada je korisničko sučelje implementirano u Reactu. Korisnicima se
omogućuje upravljanje solarnim elektranama i prognostičkim modelima te
vizualizacija i usporedba generiranih prognoza proizvodnje i informacija o
modelima. Sastoji se od web aplikacije na klijentskoj strani, API usluge na
serverskoj strani, prognostičkih modela te baze podataka za pregled i pohranu
podataka. U radu je opisana motivacija za izradu sučelja platforme, korištene
tehnologije, specifikacija zahtjeva i funkcijska specifikacija, dizajn arhitekture te
pregled korištenja sučelja platforme.This master's thesis focuses on the development of a platform for forecasting the
production of solar power plants in Croatia, with the aim of improving efficiency and
optimizing the management of solar power plants. The platform is developed using
modern web technologies, with the main focus on the user interface implemented in
React. It enables users to manage solar power plants and forecasting models, as
well as to visualize and compare generated forecasts and model information. The
platform consists of a web application on the client side, an API service on the server
side, forecasting models and a database for data review and storage. The thesis
describes the motivation for developing the platform's interface, the technologies
used, the requirements specification and functional specification, the architecture
design and an overview of the platform's interface
Computational graph isomorphism testing
No full textTema ovog rada je algoritam koji rješava problem izomorfnosti grafova. Razvijena je mobilna aplikacija koja implementira algoritam za provjeru izomorfnosti grafova temeljen na izračunu certifikata grafa. Certifikat grafa je jedinstveni potpis koji omogućuje provjeru izomorfnosti između dva grafa. Postupak za izračun certifikata se dijeli na dva algoritma, jedan prilagođen stablima koji ima linearnu vremensku složenost i drugi za općenite grafove, čija je vremenska složenost neodređena. Aplikacija omogućuje unos dvaju grafova s najviše osam vrhova, pri čemu se grafovi mogu unijeti grafički ili pomoću matrica susjedstva. Program bira odgovarajući algoritam za izračun certifikata, a zatim usporedbom certifikata dvaju grafova javlja korisniku jesu li oni izomorfni. U radu su popisani svi neizomorfni grafovi s pet vrhova i njihovi certifikati. Analizirana je i učinkovitost algoritama te rezultati pokazuju kako na vrijeme izvršavanja utječu veličina i struktura grafova.The topic of this thesis is an algorithm that solves the problem of determining graph isomorphism. A mobile application was developed that implements an algorithm for graph isomorphism verification based on graph certificate calculation. Graph certificate is a unique signature that enables the verification of isomorphism between two graphs. The certificate calculation process is divided into two algorithms, one for trees, which has linear time complexity, and another for general graphs, whose time complexity is undefined. The application allows the input of two graphs with up to eight vertices, which can be entered either graphically or using adjacency matrices. The program selects the appropriate algorithm for certificate calculation, compares the certificates of the two graphs and informs the user whether they are isomorphic. The paper also lists all non-isomorphic graphs with five vertices and their certificates. Additionally, the efficiency of the algorithms was analyzed and the results show that execution time is significantly influenced by the size and structure of the graphs
Reducing communication overhead in the IoT–edge–cloud continuum: A survey on protocols and data reduction strategies
No full textThe adoption of the Internet of Things (IoT) deployments has led to a sharp increase in network traffic as a vast number of IoT devices communicate with each other and with IoT services in the IoT–edge–cloud continuum. This increase in network traffic poses a major challenge to the global communications infrastructure since it hinders communication performance and also puts significant strain on the energy consumption of IoT devices. To address these issues, efficient and collaborative IoT solutions which enable information exchange while reducing the size of transmitted data and associated network traffic are crucial. This survey provides a comprehensive overview of the communication technologies and protocols as well as data reduction strategies that contribute to this goal. First, we present a comparative analysis of prevalent communication technologies in the IoT domain, highlighting their unique characteristics and exploring the potential for protocol composition and joint usage to enhance overall communication efficiency within the IoT–edge–cloud continuum. Next, we explore various data traffic reduction techniques tailored to the IoT–edge–cloud context and evaluate their applicability and effectiveness on resource-constrained end devices. Finally, we investigate the emerging concepts that have the potential to further reduce communication overhead in the IoT–edge–cloud continuum, including cross-layer optimization strategies and Edge AI techniques for IoT data reduction. The paper offers a comprehensive roadmap for developing efficient and scalable solutions across the layers of the IoT–edge–cloud continuum that are beneficial for real-time processing to alleviate network congestion in complex IoT environments