Presented papers in English will be submitted for inclusion in the IEEE Xplore Digital Library.

Event program

Thursday, 5/23/2019 10:15 AM - 1:00 PM, Camelia 2, Grand hotel Adriatic, Opatija Papers  1. K. Burnik, D. Bjelobrk Knežević (Zagreb University of Applied Sciences, Zagreb, Croatia) Using Convolutional Neural Networks for Emoticon Classification  The problem of automatic classification of photographs, graphics and spatial data is more and more commonly addressed using advanced machine learning methods and related models like Convolutional Artificial Neural Networks (CANN). In the scope of this work, we conducted a study of application of the CANN to an example of emoticon classification. In this paper, chosen methods are described and solutions for data collection and processing, training and evaluation of the model are offered. The presented results show successful application of the model. 2. K. Babić, A. Meštrović (Sveučilište u Rijeci, Rijeka, Croatia) Visualizations of the Training Process of Neural Networks  In this paper we present an approach to the visualizations of the neural networks training process. The main goal of such visualizations is to better understand the training process of neural networks, and to help with the creation and optimization of new neural networks. We implemented a tool which can be used to create a large number of visualizations by using different combinations of created transformations (for example: diff, plot2, angle-path, vec-dist-dev). To be able to visualize the learning process, we had to record all the weights from every nth iteration. For experiments we used three neural networks: a neural network for digit recognition (DR), a neural network for word2vec representation of words (WV), and a neural network which learned the simple logical gate XOR, (XR). For the purpose of further comparisons, we also constructed one example of a damaged network (DR', WV', XR') for each of the three original networks (DR, WV, XR). We constructed over 30 visualizations and then chose 6 visualizations which could be easily interpreted and present them in this paper. The results indicate that there may exist some regularities in the visualizations of original networks and that damaged network visualizations differ from the original networks. 3. M. Bizjak, P. Peer, Ž. Emeršič (University of Ljubljana, Faculty of Computer and Information Science, Ljubljana, Slovenia) Mask R-CNN for Ear Detection  Ear detection is an important step in ear recognition pipeline as it makes or breaks the system. However, in the literature there is arguably the lack of ear detection approaches available. This poses a problem for opening ear recognition system to wider use and applications in commercial systems. To tackle this problem we present the use of Mask R-CNN for pixel-wise ear detection. Furthermore, we directly compare our approach to one of the previous best performing pixel-wise ear detection approach by using the same dataset and protocol. Our results with intersection over union score of 79.24% on AWE dataset show the superiority of our approach and present a viable approach for future use in ear recognition pipelines. 4. I. Permozer, T. Orehovački (Juraj Dobrila University of Pula, Pula, Croatia) Utilizing Apple’s ARKit 2.0 for Augmented Reality Application Development   When it comes to practical augmented reality applications, mobile platform tools are the most deserving. Thanks to the nature of mobile devices and their everyday usage, the ideal basis for this kind of content has inadvertently formed itself. Consequently, within the iOS development environment, Apple's Xcode program enables application development using the ARKit library which delivers a host of benefits. Among the plethora of advantages, the ARKit library introduces features such as the ability to measure distances between two points in space, horizontal and vertical plane detection, two-dimensional image recognition and the ability to detect three-dimensional objects and utilize them as triggers for displaying superimposed computer generated content on iOS 11 and later iterations of Apple's mobile operating system. 5. A. Salihbašić, T. Orehovački (Juraj Dobrila University of Pula, Pula, Croatia) Development of Android Application for Gender, Age, and Face Recognition Using OpenCV  The idea behind the face recognition system is the fact that every individual has a unique face. Like the fingerprint, an individual's face has many unique structures and features. The automatic face recognition system is based on the symmetry of the face. Facial authentication and facial recognition are challenging tasks. For facial recognition systems to be reliable, they must work with great precision and accuracy. If possible, several images of the same individual should be included in the database. Images captured taking into account different facial expressions or lighting conditions allow greater precision and accuracy of the system compared to a case where only one image of each individual is stored in the database. The face recognition method handles the captured image and compares it to the images stored in the database. If a matching template is found, an individual is identified. Otherwise, the person is reported as unidentified. This paper describes and explains in detail the entire process of developing Android mobile application for recognizing person’s gender, age and face. Face detection and recognition methods that have been used are described and explained as well as development tools used in the development of Android mobile application. The program solution describes the details of using the OpenCV library and shows the actual results of the mobile application through the images. 6. A. Crnjak, D. Jaksic, P. Poscic (University of Rijeka, Department of Informatics, Rijeka, Croatia) Query Optimization in Relational Database Systems  Query optimization is an important component in relational systems and has gained great importance over the last few years. Sometimes, the reason for a slow query response time is the system load, but the more common one is that the query is written in a way that can not be performed efficiently. Query optimization requires a series of actions and procedures to get the desired result. The aim of this paper is to present and describe some of the more familiar and prominent methods and techniques of query optimization in relational database systems. 7. I. Puja, P. Poscic, D. Jaksic (University of Rijeka, Department of Informatics, Rijeka, Croatia) Overview and Comparison of Several Relational Database Modelling Metodologies and Notations  The purpose of this work is to introduce and compare several database modelling methodologies and notations and to show their application through an example. All the concepts important for this work are defined and so are the methodologies and notations that are being compared. Moreover, a database model is shown for all five notations: IDEF1X, Crow’s foot, Chen, UML, and MIRIS. 8. A. Šag, T. Orehovački (Juraj Dobrila University of Pula, Pula, Croatia) Development of 2D Game with Construct 2  The process of developing a game represents a big challenge since it requires much more knowledge and experience in a different fields than just programming. Optimization is a key point for obtaining process workflow as well as for connecting a graphical part (GUI) with a technical side (code) of the game. In the case of Construct 2, the process itself is a very user-friendly and does not require a classical way of programming. Using the drag and drop method, we can implement visual materials into our game, and the programming itself is reduced to the events option which helps us to assign actions through selected graphics icons. Application instant preview allows fast prototyping and iterative development which enables testing during the entire creation process thus facilitating detection and resolution of problems. It also provides flexibility and the ability to see how the prototype of each version actually responds to certain actions, conditions, and events. With a good organization and understanding of powerful event system, flexible behaviors, stunning visual effects and other capabilities provided by the Construct application package, it is possible to build any conceivable 2D project in a procedural way. In this paper the process of creating 2D game by means of Construct 2 environment is thoroughly explained. 9. K. Pavlinić, L. Skorin-Kapov (Fakultet elektrotehnike i računarstva, Sveučilište u Zagrebu, Zagreb, Croatia) Network and Service Performance Case Study in a High School Equipped in the Scope of the Pilot Project e-Schools  The Croatian Academic and Research Network (CARNet) is currently coordinating a program referred to as e-Schools, aimed to strengthen the capacities of elementary and secondary education by introducing ICT into the school system. The aim of this paper is to analyze the network infrastructure deployed in the e-Schools pilot project, focusing on one target high school in the city of Varaždin as a case study to help develop an insightful methodology in the future performance analysis of the infrastructure deployed in schools. A small-scale analysis of the school network infrastructure was carried out, with an additional examination of Quality of Experience (QoE) across three scenarios that simulate the usual behavior of students and teachers, while also observing the behavior of the network infrastructure through Meraki dashboard system. Based on our findings, we concluded that students are generally satisfied with the installed network infrastructure. However, there are some problems that occur (e.g., speed problems and Internet connection problems), which need to be investigated in more detail and resolved, so as to maximize end user QoE. The presented lessons learned will help develop appropriate test methodology to be applied to schools taking part in the e-Schools program for detailed performance evaluation purposes. 10. I. Pauletić, L. Načinović Prskalo, M. Brkić Bakarić (Sveučilište u Rijeci, Rijeka, Croatia) An Overview of Clustering Models with an Application to Document Clustering  This paper presents an overview of selected clustering models and shows an application of K-Means algorithm to document clustering. In the introductory part, the definitions of basic concepts and common characteristics of clustering models are described. Then an overview of clustering models is given. The methods of clustering, basic characteristics, visualization and possible input data for each algorithm are presented. The authors also explain the assessment of each algorithm taking into consideration measures such as Rand index, homogeneity, completeness, V-measure and Silhouette coefficient. Furthermore, the paper describes the application of the K-Means algorithm to document clustering showing the final result and elaborating the procedures applied when clustering the documents. 11. L. Amelio (Faculty of Medicine and Surgery, University of Bologna, Bologna, Italy), A. Amelio (DIMES University of Calabria, Rende (CS), Italy) New Frontiers in Document Classification with Applications to Medical Context  This paper surveys on the recent trends of supervised document classification methods with a special focus on the medical context. In particular, the semantic document classification is considered, where some relevant general purpose algorithms are described. Then, relevant methods for semantic classification of medical documents are in the focus. The considered algorithms are categorised according to the adopted document representation or the way the methods are applied to the input. Finally, advancement of the state-of-the-art is proposed and described in the medical domain. 12. M. Šipek, B. Mihaljević, A. Radovan (RIT Croatia, Zagreb, Croatia) Exploring Aspects of Polyglot High-Performance Virtual Machine GraalVM  Contemporary software often becomes vastly complex, and we are required to use a variety of technologies and different programming languages for its development. As interoperability between programming languages could cause high overhead resulting in a performance loss, it is important to examine how a current polyglot virtual machine with a compiler written in a high-level object-oriented language deals with it. OpenJDK's Project Metropolis presented the GraalVM, an open-source, high-performance polyglot virtual machine, mostly written in Java. This paper presents GraalVM's architecture and its features; furthermore, examining how it resolves common interoperability and performance problems. GraalVM makes software ecosystem productive when combining various programming languages, for example, Java, JavaScript, C/C++, Python, Ruby, R, and others. The vital part of GraalVM is the Graal compiler written in Java, which allows developers to maintain and optimize code faster, simpler, and more efficient, in comparison to traditional compilers in C/C++ languages. Graal can be used as a just-in-time (JIT) or as static, ahead-of-time (AOT) compiler. Graal is an aggressively optimizing compiler implementing common compiler optimizations, with emphasis on outstanding inlining and escape analysis algorithms. This paper compares Graal with some of the best-specialized competitors, and presents our results tested within an academic environment. 13. P. Pufek, H. Grgić, B. Mihaljević (RIT Croatia, Zagreb, Croatia) Analysis of Garbage Collection Algorithms and Memory Management in Java  Significant elements of the Java Virtual Machine (JVM), as a part of the Java Platform, Standard Edition (Java SE), crucial for automatic memory management are various Garbage Collection (GC) algorithms. Since implementation of the Java Development Kit (JDK) is continuously being improved, it is difficult to ignore several new Java Enhancement Proposals (JEPs) correlated to memory management in JVM. Several different garbage collectors are implemented in addition to the existing aged Serial, Parallel, and Concurrent Mark & Sweep (CMS) GC algorithms, as well as newer Garbage-First (G1) GC. The major progress since JDK 10 is making Parallel Full GC for G1, as a default multi-threaded GC when performing collections on an entire heap. Redundant overheads could appear when GC algorithms perform garbage collection too frequently, or a too large amount of memory could be allocated when garbage is not collected regularly. The goal of new algorithms' features is optimizing the overall process of releasing space so that pause times do not affect applications' performances negatively. This research paper explores several garbage collectors available in JDK 11 by using selected benchmarking applications of the DaCapo suite for comparison of the number of algorithms’ iterations and the duration of the collection time.

Matej Janjić (Croatia), Dubravko Sabolić (Croatia), Nikola Filip Fijan (Croatia)

The discount doesn't apply to PhD students.

Matej Janjic
KSET
Unska 3
HR-10000 Zagreb, Croatia

GSM: +385 91 187 1140
E-mail: matej.janjic@fer.hr

The best papers will get a special award.
Accepted papers will be published in the ISSN registered conference proceedings. Papers written in English and presented at the Conference will be submitted for posting to IEEE Xplore.
There is a possibility that the selected scientific papers with some further modification and refinement are being published in the Journal of Computing and Information Technology (CIT).

Opatija, with its 170 years long tourist tradition, is the leading seaside resort of the Eastern Adriatic and one of the most famous tourist destinations on the Mediterranean. With its aristocratic architecture and style Opatija has been attracting renowned artists, politicians, kings, scientists, sportsmen as well as business people, bankers, managers for more than 170 years.

The tourist offering of Opatija includes a vast number of hotels, excellent restaurants, entertainment venues, art festivals, superb modern and classical music concerts, beaches and swimming pools and is able to provide the perfect response to all demands.

Opatija, the Queen of the Adriatic, is also one of the most prominent congress cities on the Mediterranean, particularly important for its international ICT conventions MIPRO that have been held in Opatija since 1979 gathering more than a thousand participants from more than forty countries. These conventions promote Opatija as the most desirable technological, business, educational and scientific center in Southeast Europe and the European Union in general.

For more details please look at www.opatija.hr/ and www.visitopatija.com.