| 09:00-13:00 Radovi
|S. Sosinskaya, R. Dorofeev, A. Dorofeev, T. Usenko (Irkutsk National Research Technical University, Irkutsk, Russian Federation)
Automation of a Decision Tree Conversion into a Fuzzy Inference System Using ANTLR
The paper discusses techniques of processing a sample of sets of numerical features of observations that relate to a certain subject area and belong to certain classes. Such techniques include well-known methods of constructing a decision tree and a fuzzy inference system. An isomorphism of the decision trees and a corresponding fuzzy inference system rule set is being justified. Algorithms of both methods are described in special languages. An approach to automated conversion of the decision tree to a fuzzy inference system using ANTLR, a tool for creating compilers, is proposed. The toolkit used, along with the creation of classes for lexical and parsing of a description in one language, allows generate a class for converting text from one language to another. The relevance of the approach is that with representing fuzzy classifying knowledge allows one to implement an expert system, allowing domain specialists to classify objects. Usage of the decision trees in expert systems is problematic. An example of applying this approach to a classification of leaf specimens originating from different plant species is given.
|P. Choudhary (National Institute of Technology Hamirpur, Hamirpur, India), P. Thota (National Institute of Technology Manipur, Imphal, India)
A Randomized Load Balancing Criteria Using Traffic Flow in SDN
Software Defined Networks (SDN) is a trending technology in the computer science field, overriding traditional computing. It gives full-fledged control over the network and administrators to manage and update the respective network according to its needs by operating the control plane of the network. SDN helps to archive a global view of the network and the data abstraction. Today's drastically increasing of network suffering from the problems of overutilization of few resources of the network. Therefore, the collisions and bottleneck bandwidth of the network is growing and hence, leading to the issue of degradation of performance and throughput. All these issues are occurring due to improper load balancing schemes and algorithms used to control the load balancing. In this paper, we are presenting the algorithm named Pathfinder, a randomized pathfinding algorithm similar to the breadth-first search (BFS) algorithm but unlike BFS, it is increasing the performance and link utilization of the system. It is dynamic traffic monitoring, and balancing algorithm will shift flow to the path which has less traffic and makes sure that almost all of the resources equally utilized
|P. Georgieva, E. Nikolova, D. Orozova (Burgas Free University, Burgas, Bulgaria)
Data Cleaning Techniques in Detecting Tendencies in Software Engineering
The world of software engineering is dynamically changing over the last decade. Providing adequate university education is one of the key goals of the academic community for providing advanced and up-to-date students' training. One direction in achieving this goal is to constantly monitor the trends in IT sector. A reliable source of information is the data from the annual survey on technology and programming languages, as well as on preferred learning methods and ways to enhance competencies, conducted amongst Stack Overflow users since 2011. In processing the data from the survey, the authors have faced several problems which have provoked interest in the more general data problem – data quality and data cleaning.
This paper looks into data quality, tools for data cleaning and the characteristics of high quality data. A classification of data problems is proposed in the context of analyzing the information about software developers. In addition, the proposed process of data cleaning in illustrated with data for 2018 and 2019.
|I. Dunđer, S. Seljan, M. Pavlovski (Filozofski fakultet Sveučilišta u Zagrebu, Zagreb, Croatia)
Automatic Machine Translation of Poetry and a Low-Resource Language Pair
Automatic machine translation is gaining more and more attention in a particular part of the research community that treats various topics from artificial intelligence, natural language processing, machine learning and data science. Machine translation, in general, could be implemented in higher education and academic curricula in a variety of possible fields and applications. It is a complex task in which a computer is utilised for the purpose of translating from source to one or more target languages without human involvement, or with a minimum of interventions. Although there are several approaches to machine translation, two are dominant today – statistical and neural machine translation. Both methods in form of two online machine translation systems are being used in this research. The aim of this paper is to examine the usability of machine translation application for poetry and a low-resource language pair, such as Croatian-German. The authors chose to use a data set that contained the works of a relevant contemporary poet of the Croatian language and the translations of his poems in German that were conducted by two professional literary translators. The paper shows the effectiveness of machine translation of poetry with regard to specialised automatic quality metrics.
|S. Seljan, I. Dunđer, M. Pavlovski (Filozofski fakultet Sveučilišta u Zagrebu, Zagreb, Croatia)
Human Quality Evaluation of Machine-Translated Poetry
The quality of literary translation was from the beginning of literacy an important factor in publishing and, as a consequence, in research and education. The quality of literary text translation is of utmost significance for researchers and students, especially in higher education. Only complete and high-standard translations are believed to be necessary for the use in the evaluation and study of style and concepts of a given author or a literary genre. This quality verification applies even more to machine translation in general, due to the fact that such translations are deemed subpar and unsuitable for further dissemination and examination. The need for human quality evaluation of machine-translated text is therefore highly emphesised, since human translations are considered to be the “gold standard” and reference translations in the machine translation process. The aim of this paper is to explore, on the example of a data set consisting of poems written by a relevant contemporary Croatian poet, the effectiveness of applying machine translation on the Croatian-German language pair in the domain of poetry, with regard to human judgment of machine translation quality. Human evaluation in this paper is conducted by taking into account two machine translation quality criteria – adequacy and fluency, after which an inter-rater agreement analysis is performed.
|V. Kondratiev, I. Otpuschennikov, A. Semenov (Matrosov Institute for System Dynamics and Control Theory of Siberian Branch of Russian Academy of S, Irkutsk, Russian Federation)
Using Decision Diagrams of Special Kind for Compactification of Conflict Data Bases Generated by CDCL SAT Solvers
In the paper we propose new algorithms for constructing compact representations of databases of conflict clauses accumulated by state-of-the-art CDCL SAT solvers.
These algorithms use the Decision Diagrams of a special kind. We consider several families of hard SAT instances and use them to compare the implementations of
the proposed algorithms and the well-known CUDD package that uses Zero-Suppressed Binary Decision Diagrams (ZBDD) for solving similar problems.
The computational experiments clearly show that our algorithm has better effectiveness compared to CUDD.
|S. Sambolek (High school Tina Ujevića, Kutina, Croatia), M. Ivašić-Kos (Odjel za informatiku Sveučilište u Rijeci, RIJEKA, Croatia)
Detecting Objects in Drone Imagery: a Brief Overview of Recent Progress
Detecting objects on unmanned Aircraft Systems (Drones) imagery is a challenging task and an under-researched problem that has been lately receiving more and more attention in the research community. When shooting with a drone, not only weather and light conditions change, but the shooting height and angle change as well during shooting, as opposed to fixed camera shots.
This paper aims to describe the possibility of using drones for search and rescue operations and to provide a comprehensive overview of the area related to the detection of persons in drone-recorded images. The paper includes a description of publicly available datasets and a comparison of state-of-the-art models used to detect persons in drone imagery and concludes with a suggestion of direction for future research.
|T. Bronzin, B. Prole, A. Stipić (CITUS, Zagreb, Croatia), K. Pap (University of Zagreb, Faculty of Graphic Arts, Zagreb, Croatia)
Individualization of Anonymous Identities Using Artificial Intelligence (AI)
Individualization of anonymous identities using artificial intelligence - enables innovative human-computer interaction through the personalization of communication which is, at the same time, individual and anonymous. This paper presents possible approach for individualization of anonymous identities in real time. It uses computer vision and artificial intelligence to automatically detect and recognize person’s age group, gender, human body measures, proportions and other specific personal characteristics. Collected data constitutes the so-called person's biometric footprint and are linked to a unique (but anonymous) identity that is recorded in the computer system, along with other information that make up the profile of the person. Identity anonymization is achieved by asymmetric encryption of the biometric footprint, with no additional personal information being stored, and integrity is ensured using blockchain technology. By using an anonymous but individualized identity, a person "unlocks" access to various content/capabilities of IT system that is accessed only through authentication. Collected data is GDPR compliant.
|I. Tomicic, P. Grd (Fakultet organizacije i informatike, Varazdin, Croatia)
Towards the Open Ontology for IoT Ecosystem’s Security
The IoT ecosystem is complex and includes hardware devices, services, software, connectivity, standards and protocols. IoT is continuously and persistently contributing to the ubiquity of technology in all aspects of human endeavours, and trends seem to indicate that myriad of IoT solutions are developed and/or bought for various purposes, from the personal home to industrial usages, with the alarming lack of security concerns, giving priority mostly to the fast and cost-effective development. Given the complex nature of such solutions, it is rarely seen that the security aspect is considered through all of the IoT ecosystem elements; also, one of the significant reasons being that the IoT solutions are not always developed in a holistic nature, within the same company, but only provide parts of the complete IoT solution. In order to enable easier and more secure development of IoT solutions, this paper presents first steps towards developing the open ontology for IoT ecosystem elements security, relating them with existing security concepts, primitives, threats, vulnerabilities and practices.
|I. Šulekić (Tehničko veleučilište u Zagrebu, Zagreb, Croatia), D. Milinković (Ekonerg d.o.o., Zagreb, Croatia), T. Špoljarić (Tehničko veleučilište u Zagrebu, Zagreb, Croatia)
Decision Tree Algorithm for Control Of Compressor Multiset in Refrigeration Industry
This paper describes novel control algorithm for the multi-compressor set. Algorithm has input parameters that are collected from each compressor in the set. According to these parameters, algorithm decides which combination of compressors is the best for optimum performance of the entire set.
|N. Stojanova, R. Vignjevikj, A. Naumoski (Ss. Cyril and Methodius University – Skopje, Faculty of Computer Science and Engineeing, Skopje, Macedonia)
GIS Analysis of Basketball Courts and Healthy Stores Relationship for Young Population in the City of Skopje
Today we live in a world where everything that we need we can obtain from the comfort of our homes. Although this is very convenient, our physical and mental health are paying the price. Physical exercises are one of the key elements in our daily lives that keep our body in top shape. Therefore, in this paper we address two key health components: exercises through sport and healthy eating. For this purpose, we have mapped most of the open fields and closed basketball courts in the city of Skopje, the population numbers of young people between the age of 5 and 25, as well as the healthy food and pharmacy stores using Geographic Information Systems. Combing the information from these three sources, we obtain data fusion spatial analysis map that we use to find the best location for opening new basketball court, and in the same time providing crucial knowledge of spatial regions where young people can have healthy food or buy medicines in a case of injury.
|R. Šajina, N. Tanković, D. Etinger (Juraj Dobrila University of Pula, Faculty of Informatics, Pula, Croatia, Pula, Croatia)
Decentralized Trustless Gossip Training of Deep Neural Networks
Novel machine learning techniques apply decentralized model training in order to mitigate data volume and privacy issues. Current approaches assume (a) node performance homogeneity, and (b) simultaneous training. These assumptions also imply that the predictive performance of the distributed models evolves uniformly. A different approach is required since a distributed decentralized network is heterogeneous and nonstationary: nodes can join or leave the network at any point in time (churn). We propose a novel protocol for exchanging the model knowledge between peers using a gossip algorithm combined with the stochastic gradient descent (SGD). Our method has the advantage of being fully asynchronous, decentralized, trustless, and independent of the network size and the churn ratio. We validated the proposed algorithm by running network simulations in various scenarios.
|S. Delalić, A. Alihodžić (Faculty of Science, University of Sarajevo, Sarajevo, Bosnia and Herzegovina), M. Tuba (Singidunum University, Belgrade, Serbia), E. Selmanović, D. Hasić (Faculty of Science, University of Sarajevo, Sarajevo, Bosnia and Herzegovina)
Discrete Bat Algorithm for Event Planning Optimization
Many public figures, companies and associations are planning events in different cities and at the same time have active profiles on social media. The planning process requires processing a large amount of data and different parameters when choosing the best event venue. At the same time, social media captures a large number of fan actions per day. This paper describes the process of selecting the most appropriate cities for organizing an event, aided by data collected from social media. The problem is defined as a combinatorial optimization problem. A modified metaheuristic bat algorithm was proposed, implemented, and described in detail to solve the problem. Although the original bat algorithm is designed to solve continuous optimization problems, the implemented bat algorithm is adapted to solve the defined problem. The algorithm is compared to the exhaustive search method for smaller instances, and to the greedy and genetic algorithm for larger instances. The algorithm was tested on benchmark data on cities in European countries, as well as on real data collected from pages on the social network Facebook. Bat algoirtam has shown superior results compared to other techniques, both in time and in the quality of the solutions generated.
|I. Tomicic, M. Schatten (Fakultet organizacije i informatike, Varazdin, Croatia)
A Conceptual Network Analysis of Gamification Practices in Primary and Secondary Education
Gamification is not a new concept, but it has gained a significant momentum in the past years for several reasons, including the ubiquity of technology and the growing number of technologically savvy individuals who are routinely using smartphones and other computer devices for various tasks, including playing games and using e-learning platforms, but also because of the growing ease in the use of the game design applications. Gamification is mostly defined as the use of game design elements and game mechanics in a non-game context, with the main objective of engaging and interacting with users. The existing research points to the possibility of improving the performance of students in the learning process, and this paper presents a state of the art, conceptual network analysis of gamification practices in learning processes with the aim to better understand game-based learning in primary and secondary education. The results from this paper would be used for the development of an ai-based learning platform.
| 15:00-19:00 Radovi
|B. Gašperov, F. Šarić, S. Begušić, Z. Kostanjčar (Fakultet elektrotehnike i računarstva, Zagreb, Croatia)
Adaptive Rolling Window Selection for Minimum Variance Portfolio Estimation Based on Reinforcement Learning
When allocating wealth to a set of financial assets, portfolio optimization techniques are used to select optimal portfolio allocations for given investment goals. Among benchmark portfolios commonly used in modern portfolio theory, the global minimum variance portfolio is becoming increasingly popular with investors due to its relatively good performance which stems from both the low-volatility anomaly and the avoidance of the estimation of first moments i.e. mean returns. However, estimates of minimum variance portfolio weights significantly depend on the size of the rolling window used for estimation, especially considering the non-stationarity of the underlying market dynamics. In this paper, we use a model-free policy-based reinforcement learning framework in order to directly and adaptively determine the optimal size of the rolling window. Training is done on a subset of trading stocks from the NYSE. The resulting agent achieves superior performance when compared against multiple benchmarks, including those with fixed rolling window sizes.
|A. Gribl, D. Petrinović (University of Zagreb, Faculty of Electrical Engineering and Computing, Zagreb, Croatia)
Synthetic Astronomical Image Sequence Generation
Astronomical image reconstruction from multiple observations requires detailed knowledge of the imaging system, and the accuracy of the reconstruction is determined by the imaging model correctness. The parameters of the imaging model are numerous, and they must be known precisely for successful reconstruction. When these parameters are not known exactly but are estimated as a part of the processing pipeline, it is very important to understand the influence of their estimation error on the final reconstruction accuracy. The best way to measure this sensitivity is through detailed simulation of the whole imaging process that takes into account all of the model parameters, which can then be artificially manipulated in order to determine the required accuracy. This paper describes a novel approach to generate a synthetic astronomical image from the chosen star catalogs. The synthetic image is determined by external camera parameters (viewpoint - azimuth, elevation, and rotation), by exposure time, by the selected optics (focal length which specifies the view angle), by the sensor dimension and its resolution, by the lens f-number which determines the diffraction pattern on the circular aperture, by the spherical and chromatic aberrations of the lens what cause additional spreading of the object image on the sensor (PSF), by the lens vignetting, by photon sensitivity of the sensor, sensor gain (ISO) and noise, color filter array in front of the sensor, by the unwanted residual motion of the imaging system during exposure, and finally by the portion of the sky enclosed in the frame, which depends on the observation location, date, exact time and external camera and/or telescope mount parameters. Synthesis assumes point sources extracted from the chosen catalogs and described through their angular position and magnitude in different spectral bands. The model parameters are tuned to the actual raw image captured by the SLR camera and chosen lens. Generation of a series of synthetic images of approximately the same part of the sky is performed, but with small shifts (small perturbations of the external parameters of the camera) that results in a series of slightly shifted frames, in which the light field integrates into different spatially displaced sampling grids (with integer and fractional shifts), thus facilitating high-resolution image reconstruction from spatially subsampled projections.
|R. Šajina, N. Tanković, D. Etinger (Faculty of Informatics in Pula, Pula, Croatia)
Novel Class Detection in Non-stationary Streaming Environment with a Discriminative Classifier
In a data streaming environment, one of the biggest challenges for a machine learning classifier is to detect the changes in the concepts that the data corresponds to - a phenomenon called Concept Drift. It can manifest in different modes: existing classes can continually evolve, experience a sudden shift, or a novel class can emerge.
Algorithms for novel class detection using cluster-based techniques and statistical approaches applied to the model outputs rely on the assumption that the feature space of the data posses some distance metric governing the class affiliation. Therefore, the novel class will correspond to a significant distance from the known clusters.
Most of the time, these assumptions are correct, but the resulting algorithms are challenging to apply on higher-dimensional data such as images. In this paper, we present a novel approach called Discriminative Classifier Detector (DCD) for detecting concept evolution. DCD trains alongside the classification model. The primary model used to evaluate our approach is a Convolutional Neural Network (CNN) classifier for which the proposed DCD is a densely layered neural network. DCD is applied to the outputs of CNN's penultimate layer to achieve its independence against the number of output classes, which enables the underlying CNN model to evolve independently. DCD requires no structural change when the detected novel class is added to the CNN classifier. We demonstrate the effectiveness of our approach on several well-known datasets.
|A. Naumoski, G. Mirceva, K. Mitreski (Ss. Cyril and Methodius University – Skopje, Faculty of Computer Science and Engineeing, Skopje, Macedonia)
Evaluation of Diatoms Biodiversity Models by Applying Different Discretization on the Class Attribute
One of the main goals of knowledge discovery from environmental data is through data analysis to find the relationship between the living organisms, represented with the diversity of the diatoms community members, and the characteristics of the environment. This is very important information for both ecologists and decision makers. Therefore, in this paper we apply various machine learning algorithms for revealing this relationship, by using different number of discretization levels for the target attribute. The target attribute represents the biodiversity index of the community and it is calculated based on the abundances of the diatoms. For building models, different types of machine learning algorithms are considered, including decision trees, rule induction algorithms, neural networks and Naïve Bayes. The obtained models are also examined regarding resistance to over-fitting, as well as statistical significance.
|G. Mirceva, A. Naumoski, A. Kulakov (Faculty of computer science and engineering, Ss. Cyril and Methodius University in Skopje, Skopje, Macedonia)
Classifying Protein Structures by Using Protein Ray Based Descriptor, KNN and FuzzyKNN Classification Methods
Bioinformatics community continuously studies the protein molecules and the processes in which they are involved. To understanding these processes, the researches need to understand the characteristics and functions of various group of proteins. Therefore, classification of protein structures is one of the most studied problems in proteomics. Although many methods are developed for this purpose, still there is a need for fast and accurate methods that would provide classification of protein structures. In this paper, our aim is to develop a method that would provide accurate classification of proteins. First, we extract feature vectors that hold the main geometrical characteristics of the protein structures represented by their tertiary structures. For that purpose we use our protein ray based descriptor, which represents how the amino acid residues are placed with respect to the center of mass. After feature extraction, next we applied two instance learning methods (k nearest neighbors (KNN) and FuzzyKNN classification methods) for generating prediction models. Besides the KNN classifier, which is based on classical set theory, we also used the FuzzyKNN that is based of fuzzy set theory. For evaluation, we used a part of the SCOP database that contains knowledge about the domain in which a given protein belongs to. The paper presents the experimental results achieved by using the two classification methods by considering various number of nearest neighbors.
|B. Dalbelo Bašić (Faculty of Electrical Engineering and Computing – TakeLab – University of Zagreb, Zagreb, Croatia), M. di Buono (UNIOR NLP Research Group – University of Naples “L’Orientale”, Naples, Italy)
An Analysis of Early Use of Deep Learning Terms in Natural Language Processing
In this paper, we present the preliminary results on the analysis of deep learning terms used for natural language processing (NLP) tasks. We propose a statistical analysis of papers published from 2012 to 2015 in the main ACL conferences. Our aim is investigating which DL term, and consequently which DL method, is mostly used for each specific NLP task, since its introduction in the field. In order to do this, our first contribution is the development of two terminological lists, referring respectively to DL methods for text analysis and NLP tasks. The list of deep learning terms contains 41 terms and acronyms, as well as a NLP term list contains 145 terms and acronyms. From our corpus, the frequencies of various terms have been extracted with respect to the ACL conference and the publication year. After the preliminary data analysis, we decided to restrict the extraction process to abstract texts. We applied multivariate techniques called correspondence analysis in order to visualize and evaluate the joint behavior of our variables.
|I. Cherepanov, A. Mikhailov, A. Shigarov, V. Paramonov (IDCST, Irkutsk, Russian Federation)
On Automated Workflow for Fine-tuning Deep Neural Network Models for Table Detection in Document Images
Nowadays methods and software for extracting tables from document images and portable documents (PDF) continue to be actively developed. One of the promising approaches to this task is the usage of fine-tuned object detection models. However, this approach involves many manipulations with data preparation and training process configuration. This paper proposes an automated workflow for fine-tuning deep neural network models for the table detection in document images. It enables us to automate two sub-tasks: (i) preparing a training dataset in the PascalVOC format with image transformation and augmentation; (ii) training a table detection model by using the well-known Faster R-CNN architecture. Implementation of the workflow design simplifies the use of the approach proposed by decreasing the number of required manipulations.
|M. Horvat (University of Zagreb, Faculty of Electrical Engineering and Computing, Zagreb, Croatia)
StimSeqOnt: An Ontology for Formal Description of Multimedia Stimuli Sequences
Sequences of multimedia documents are successfully used in laboratory settings and in practice to deliberately elicit specific emotional reactions. To ensure a successful experiment the emotion provoking stimuli must be selected carefully and have a specific order in which they are presented to the participants. Temporal aspect – duration of individual stimuli within sequences, duration of whole sequences and pauses between stimuli and sequences – must also be chosen with great care. Construction of effective sequences is a delicate and time consuming activity which requires significant group manual effort from domain experts. To facilitate this task we propose a new ontology called StimSeqOnt for formal description of stimuli sequences. The ontology is written in OWL DL language and provides formal and sufficiently expressive representation of affective concepts, high-level semantics, stimuli documents, multimedia formats and repositories used. In StimSeqOnt all relevant metadata about stimuli sequences may be stored as formal concepts. If available, elicited physiological data of previously exposed participants are available for comparison thereby enabling prediction of emotional responses. The StimSeqOnt is designed in compliance with ontology guidelines to facilitate sharing and reuse of expert knowledge.
|Đ. Pašić, D. Kučak (Visoko učilište Algebra, Zagreb, Croatia)
Machine Learning Model for Detecting High School Students as Candidates for Drop-out from a Study Program
Transition from high school to university is not successful for all students. Before enrolling in a program, the Admission Office tries to help those students to decide whether that program is best suited for them. They do so by using collected empirical data. The challenge is to classify which students would successfully finish the program and which would not. The students who are most likely not to finish the program successfully should be warned that their decision is not necessarily the best decision for them, and they should consider some other possibilities. This paper proposes one possible criterion for the given challenge: we will develop a machine learning model with the collected data from the high schools which the students have attended prior to Algebra and calculate the probability of not finishing the program successfully. Students who would be classified by the model as unsuccessful will receive recommendation for another study program.
|M. Frković (University of Zagreb Faculty of Electrical Engineering and Computing, Zagreb, Croatia), N. Čerkez (College of Information Technologies (VSITE) , Zagreb, Croatia), B. Vrdoljak (University of Zagreb Faculty of Electrical Engineering and Computing, Zagreb, Croatia), S. Skansi (University of Zagreb Faculty of Croatian Studies, Zagreb, Croatia)
Evaluation of Structural Hyperparameters for Text Classification with LSTM Networks
In natural language processing, most problems can be interpreted as text classification tasks, which makes this type of task a central one. A natural subdivision of more complex types of text classification can be made according to whether the classification is multilabel or multiclass, where both of these can be tackled with either a multiclass classifier or with a combination of several binary classifiers. An example of the problem which offers a natural way of comparing multiclass and binary classification on the same data, which makes the results comparable, is a classification of text author MBTI personality type. The dataset used is PersonalityCafe MBTI. We focus our comparison on structural hyperparameters, which are the hyperparameters pertaining to network structure. Structural hyperparameters are necessary for specifying the network itself, which makes them a primary concern in its construction. The hyperparameters investigated in our paper are the number of hidden layers and layer size. Through a number of experiments, we demonstrate the choice of hyperparameters and conclude with general hyperparameter selection recommendations based on our results.
|A. Davydov, A. Larionov, N. Nagul (Matrosov Institute for System Dynamics and Control Theory SB RAS, Irkutsk, Russian Federation)
On Checking Controllability of Specification Languages for DES
The paper provides further development of the authors' original approach to the representation and properties checking of discrete event systems. The approach suggested is based on the automated inference in the calculus of positively constructed formulas (PCF). The discrete event system is supposed to be modeled in the form of finite automata within the framework of Ramage-Wonham supervisory control theory. It is shown how constructive inference helps to build the product of two finite automata which results in the automaton with accessible states only. Based on the nonmonotonic logical inference of PCF, a new method is presented for checking the controllability of formal languages describing specifications on the functioning of discrete event systems.
|T. Petković (University of Zagreb Faculty of Electrical Engineering and Computing, Zagreb, Croatia), S. Gasparini (ENSEEIHT, Tolouse, France), T. Pribanić (Universityof Zagreb Faculty of Electrical Engineering and Computing, Zagreb, Croatia)
A Note on Geometric Calibration of Multiple Cameras and Projectors
Geometric calibration of cameras and projectors is an essential step which must be performed before any imaging system can be used. There are many well known geometric calibration methods for calibrating systems comprised of multiple cameras, but simultaneous geometric calibration of multiple projectors and cameras is not so well discussed. This leaves unresolved many practical issues which must be considered to achieve the simplicity of use required for real world applications. In this work we discuss several important components of a real-world geometric calibration procedure which is used in our laboratory to calibrate surface imaging systems comprised of many projectors and cameras. We specifically discuss the design of the calibration object and the image processing pipeline used to analyze it in the acquired images. We also provide quantitative calibration results in the form of reprojection errors and compare them to the classic approaches such as Zhang’s calibration method.
Slobodan Ribarić (Croatia), Andrea Budin (Croatia)
Patrizio Campisi (Italy), Bojan Cukic (United States), Ivo Ipšić (Croatia), Marina Ivašić-Kos (Croatia), Zongmin Ma (China), Neeta Nain (India), Nikola Pavešić (Slovenia), Vitomir Štruc (Slovenia), Zheng-Hua Tan (Denmark)
PRIJAVA / KOTIZACIJE
CIJENA U EUR-ima
|Članovi MIPRO i IEEE
|Studenti (preddiplomski i diplomski studij) te nastavnici osnovnih i srednjih škola
Popust se ne odnosi na studente doktorskog studija.
Fakultet elektrotehnike i računarstva
10000 Zagreb, Hrvatska
Tel.:+385 1 612 99 52
Fax: +385 1 612 96 53
Ericsson Nikola Tesla d.d.
10000 Zagreb, Hrvatska
Tel.:+385 1 365 34 23
Fax: +385 1 365 3548
Najbolji radovi bit će nagrađeni.
Prihvaćeni radovi bit će objavljeni u zborniku radova s ISSN brojem. Prezentirani radovi na engleskom jeziku bit će poslani za uključenje u digitalnu bazu IEEE Xplore (iznimno će se za objavu uzeti u obzir radovi autora s opravdanim razlogom nemogućnosti prezentiranja rada).
Postoji mogućnost da se odabrani znanstveni radovi uz određenu doradu objave u međunarodnom časopisu Journal of Computing and Information Technology (CIT).
Opatija, sa 170 godina dugom turističkom tradicijom, vodeće je ljetovalište na istočnoj strani Jadrana i jedno od najpoznatijih na Mediteranu. Ovaj grad aristokratske arhitekture i stila već 170 godina privlači svjetski poznate umjetnike, političare, kraljeve, znanstvenike, sportaše, ali i poslovne ljude, bankare, menadžere i sve kojima Opatija nudi svoje brojne sadržaje.
Opatija svojim gostima nudi brojne komforne hotele, odlične restorane, zabavne sadržaje, umjetničke festivale, vrhunske koncerte ozbiljne i zabavne glazbe, uređene plaže i brojne bazene i sve što je potrebno za ugodan boravak gostiju različitih afiniteta.
U novije doba Opatija je jedan od najpoznatijih kongresnih gradova na Mediteranu, posebno prepoznatljiva po međunarodnim ICT skupovima MIPRO koji se u njoj održavaju od 1979. godine i koji redovito okupljaju preko tisuću sudionika iz četrdesetak zemalja. Ovi skupovi Opatiju promoviraju u nezaobilazan tehnološki, poslovni, obrazovni i znanstveni centar jugoistočne Europe i Europske unije općenito.
Detaljnije informacije se mogu potražiti na www.opatija.hr i www.visitopatija.com.