Pages

3D reconstruction system on smartphone
3D reconstruction system on smartphone
Marko Lelas
In order to develop a novel 3D reconstruction system on smartphone, this work first gives an analysis of the state of the art active and passive stereo 3D approaches. Drawbacks as well as the advantages of the mentioned approaches are identified and a novel approach to stereo correspondence estimation is presented as a combination of the main ideas of the approaches. In brief, the proposed approach considers image pixel values both in spatial and time domain. That way inherent downsides,...
4D medical data compression architecture
4D medical data compression architecture
Martin Žagar
This thesis presents a novel framework for four-dimensional medical data compression architecture. This framework is based on different procedures and algorithms that detect time and spatial redundancy in recorded MRI volumes. Motion in time is analyzed through motion estimation based on neural networks. Motion estimation is used to eliminate a large amount of temporal and frequency redundancies that exists in sequences of 3D data. Combination of segmentation, block matching and motion field...
A Framework for Service-level End-to-end Quality of Service Negotiation and Adaptation
A Framework for Service-level End-to-end Quality of Service Negotiation and Adaptation
Lea Skorin-Kapov
With the move towards a converged, all-IP next generation network (NGN) offering end users personalized, advanced multimedia services anywhere and anytime, a key challenge to be addressed is support for the service-level negotiation and dynamic adaptation of Quality of Service (QoS) parameters. Considering that the majority of existing approaches presented in literature and standards focus on solving particular aspects of this problem, it may be concluded that there is a need for a...
A deep learning model for estimation of human body measurements from images
A deep learning model for estimation of human body measurements from images
Kristijan Bartol
The understanding of body measurements in and between populations is important and has many applications in medicine surveying, the fashion industry, fitness, and entertainment. Recent advances in human body measurement and shape estimation have been significantly driven by statistical models and deep learning, enabling methods that estimate 3D human meshes from 3D point clouds and 2D images - so called mesh regression methods. This thesis builds upon the state-of-the-art mesh regression...
A framework for collection, contextual enrichment and advanced analytics of automotive data
A framework for collection, contextual enrichment and advanced analytics of automotive data
Hrvoje Vdović
The thesis addresses the challenges related to collection, contextual enrichment and analytics of automotive data. Current trends that focus on vehicle connectivity, i.e. vehicle-2-infrastructure and vehicle-2-vehicle, are seen as beneficial for data collection and analysis. However, collecting data from connected vehicles is not a simple task because connected vehicle hardware and software are a closed system meant to be used by vehicle manufacturers only. Additionally, when automotive data...
A framework for end-to-end qualiti of service analysis in a multi-provider environment
A framework for end-to-end qualiti of service analysis in a multi-provider environment
Irena Grgić Gjerde
The structure of today’s telecommunications market is getting more complex than ever due to multiple reasons. First of all, the traditional monopolistic Public Network Operators’ (PNOs) market is moving towards an open market with many various players. Such liberalisation is followed by a strengthening of regulatory directives that support and favour the open market, e.g. Open Network Provision (ONP) directives from the European Commission (EC). Other characteristics of such market...
A geometric approach for generating feasible configurations of robotic manipulators
A geometric approach for generating feasible configurations of robotic manipulators
Filip Marić
Most robotic manipulators, and especially those designed with autonomous operation in mind, consist of a series of joints that rotate about a single axis, also known as revolute joints. These mechanisms give robotic manipulators the degrees of freedom and versatility similar to that of the human arm, which they are designed to outperform. However, this results in a geometry of motion or kinematics that makes all aspects of robotic manipulation challenging from a computational perspective. A...
A hyper-heuristic approach to achieving long-term autonomy in a heterogeneous swarm of marine robots
A hyper-heuristic approach to achieving long-term autonomy in a heterogeneous swarm of marine robots
Anja Babić
For the purpose of enabling long-term autonomy of a heterogeneous swarm of marine robots, task allocation and sequencing were introduced into the system’s energy management procedures and agent interactions. In a scenario where the system needs to autonomously go about its monitoring mission and survive long-term, the available maximum capacity of five surface vehicles - aPad platforms which represent the charging hubs of the system - is usually outnumbered by the number of active charging...
A probabilistic quality of service model for nondeterministic service compositions.
A probabilistic quality of service model for nondeterministic service compositions.
Adrian Satja Kurdija
Cloud computing becomes the prevailing aspect of software engineering. In Service-Oriented Architecture (SOA), various atomic services respond to various client requests in a cloud. Selecting the actual service instance to process a particular request can be an issue when requirements for multiple Quality of Service (QoS) properties need to be satisfied for many users simultaneously. The problem becomes more complex if we take into account the compositeness of users’ applications, which...
Active RC Filter Design Using Impedance Tapering
Active RC Filter Design Using Impedance Tapering
Dražen Jurišić
The design procedure of low-sensitivity, low-pass (LP) 2nd- and 3rd-order class-4 Sallen and Key active resistance-capacitance (RC) allpole filters, using impedance tapering, has already been published [1]. There a ladder structure in the positive feedback loop of an operational amplifier was used. In this work, desensitisation using impedance tapering is applied to LP, HP and BP filters up to 6th-order, with single operational amplifier. Realizability constraints are also given. It was...
Adaptive Instantaneous Frequency Estimation of Noisy Nonstationary Signals
Adaptive Instantaneous Frequency Estimation of Noisy Nonstationary Signals
Jonatan Lerga
Methods for a noisy nonstationary signal instantaneous frequency (IF) estimation from its time-frequency representation (TFR) have been proposed. The IF estimation error exhibits bias and variance dependant on the TFR analysis window width. Due to the opposite tendency of bias and variance with respect to the lag window width, a proper width has been found as the one providing the bias-to-variance tradeoff that reduces the estimation error. The adaptive window widths have been obtained by...
Adaptive compressive sensing in polynomial and shift-invariant spaces
Adaptive compressive sensing in polynomial and shift-invariant spaces
Tin Vlašić
This thesis concerns solving inverse problems, i.e., the problem of reconstructing an unknown signal, image, or shape from a set of measurements. The measurements are obtained from the unknown data by a forward process, which is typically known and non-invertible. Amongst the numerous tasks that fit under this framework, the thesis focuses particularly on compressive sensing (CS) and the inverse obstacle scattering problem. Compressive sensing is a sampling and reconstruction technique that...

Pages