Postgraduate doctoral studies on Institute of informatics, Slovak academy of sciences

Institute of Informatics SAS is the accredited external educational institution for postgraduate doctoral studies in the study program  “Applied informatics” 9.2.9 (the agreement with FIIT STU). Deadline for PhD study registration is May 31, 2019.

PhD thesis for academic year 2019/2020:

study program  9.2.9 Applied informatics,
Theme: Supervisor: Annotacion:
New Methods for Large-scale Real-time Collection, Aggregation and Processing of Geo-mapped Data Balogh Z. (annotation)
Fast algorithms for the computation of perfect reconstruction filter banks (defined as the block transforms) employed in the international audio coding standards. Britaňák V. (annotation)
Bio-inspired optimization methods Budinská I. (annotation)
Inteligen environmnets and IoT Budinská I. (annotation)
Parallelization of computer simulation of fires Glasa J. (annotation)
Use of high performance computing for modeling of fire spread Glasa J. (annotation)
Computer modeling of flows in road tunnel in fire conditions Glasa J. (annotation)
Distributed large data processing Hluchý L. (annotation)
Artificial Intelligence Methods in Cyber Security Hluchý L. (annotation)
Deep neural networks for applications in image processing and computer vision Malík P. (annotation)
Applied soft computing for complex solutions Nguyen G. (annotation)
Sentiment analysis in text Nguyen G. (annotation)
FURHAT Robotics’s Robotic Head as an Expressive Voice Assistant input/output device in Slovak Rusko M. (annotation)
Automatic information retrieval using the natural language processing Rusko M. (annotation)
Deep neural networks in end-to-end speech recognition Rusko M. (annotation)
Automatic measurement of stress from voice Rusko M. (annotation)
Application development, deployment and orchestration in Cloud computing Dinh Viet Tran (annotation)

Studying at the Institute of Informatics SAS has the following benefits:

  • Working with up-to-date technologies and access to a powerful computing cluster of the II SAS.
  • Involvement in research projects at national and international levels.
  • Cooperation in production of interesting applications for practice.
  • International mobility, financial coverage of travel expenses for active participations in international conferences.
  • Ph.D. internship at foreign scientific institutions.
  • The possibility of accommodation in a boarding house of SAS (in Devinska Nova Ves, 15 min. by bus from II SAS) in renovated apartments for a good price (75 € / month).
  • A scholarship for the academic year 2019/2020 was
    a) scholarship at least € 561.50 (before completing a state degree exam) + possibility of a surcharge when working on projects,
    b) scholarship at least € 646.50 (after completing a state degree exam) + possibility of a surcharge when working on projects.
  • Bonus for a successful Dissertation thesis defense.

Other employee benefits

  • flexible work schedule, weekly working time of 37,5 hours
  • additional 5 paid vacation days
  • advantageous reimbursement of sick leave
  • bonuses for publications and year-end bonuses
  • bonuses or incentives for publications (according evaluation by the Scientific board of II SAS)
  • Recreational options – holiday stay in the Congress centre in Smolenice, the Academia Congress centre in Stará Lesná, and an institutional owned cottage in Vyhne
  • flexible working time
  • teambuilding activities (an institutional field day e.g. in Smolenice castle)

If you have questions or would you like to work with us, contact us:

 

PhD students – 2017/2018

Applied informatics
PhD student: Supervisor:
Ing. Ondrej Kachman Doc. Ing. L. Hluchý, CSc., konzultant špecialista: Ing. M. Baláž, PhD.
RNDr. Andrej Ridzik Doc. Ing. L. Hluchý, CSc., konzultant špecialista: Ing. M. Baláž, PhD.
Cybernetics
PhD student: Supervisor:
Ing. Kasanický Tomáš Ing. I. Budinská,PhD.
Ing. Juraj Pristach Ing. I. Budinská, PhD.

Anotácie jednotlivých tém:

New Methods for Large-scale Real-time Collection, Aggregation and Processing of Geo-mapped Data
Supervisor: Ing. Zoltán Balogh, PhD.
Department: Department of Parallel and Distributed Information Processing
The research is focusing on the topic of large-scale collection of geographically mapped structured or semi-structured data from a large number of sensors. Collection of data occurs from mobile computing devices such as smart mobile phones, smart watches or single-board computers. Research can also tackle topics such as contextual real-time notification, semantic data modelling, continuous visualization of dynamically aggregated information or a design of a novel distributed architecture for credible and secure real-time data collection using specialized data aggregation points in the network. There is an opportunity to exploit the research results in projects in the domains of crisis management support or intelligent transportation control.

(Back)

Fast algorithms for the computation of perfect reconstruction filter banks (defined as the block transforms) employed in the international audio coding standards.
Supervisor: RNDr. Vladimír Britaňák, DrSc
Department: Department of Parallel Computational Methods and Algorithms
Perfect reconstruction filter banks defined as the block transforms are basic processing components in the international audio coding standards such as, MPEG-1/2 layer III known as MP3, MPEG-4 Advanced Audio Coding (AAC), MPEG-4 High Efficiency AAC (HE-AAC), MPEG-4 AAC Enhaced Low Delay (AAC-ELD), and proprietary audio coding algorithms such as, Sony ATRAC/ATRAC2/SDDS and Dolby Digital AC-3/Dolby Digital Plus. Modified discrete cosine/sine transform (MDCT/MDST), modulated lapped transform (MLT), modulated complex lapped transform (MCLT), and various exponential and real-valued cosine-modulated QMF (Quadrature Mirror Filters) banks are very time-consuming operations, in particular for block sizes 256, 512, 640, 1024, 1920, 2048 and 4096. Therefore, the design of fast algorithms in terms of the minimal computational complexity for their real-time hardware/software implementation is of great importance (main motivation of the research).
The topic involves the following separate research sub-topics (mathematical properties of perfect reconstruction filter banks in the time and frequency domain; design and implementation of perfect reconstruction filter banks; matrix representations perfect reconstruction filter banks; integer approximation of perfect reconstruction filter banks).

(Back)

Bio-inspired optimization methods
Supervisor: Ing. Ivana Budinská, PhD.
Department: Department of Modelling and Control of Discrete Processes
Bio-inspired methods belong to a huge and diverse set of algorithms that are often used in practice to solve complex optimization problems when a problem-specific algorithm cannot be efficiently used. The topic is focused on the research and development of nature-inspired algorithms for solving one-parameter optimization with multiple decision variables and on the development of algorithms for multivariable optimization and modeling of such optimization problems. Such problems occur in many areas of science, technology, economics and logistics. The optimal solution must be found in compromise between two or more conflicting optimization goals. The following methods and algorithms will be used: Evolutionary algorithms, particle swarm optimization, ant colony algorithm and machine learning.

(Back)

Inteligen environmnets and IoT
Supervisor: Ing. Ivana Budinská, PhD.
Department: Department of Modelling and Control of Discrete Processes
Intelligent environment concept propose a vision of a society with effective knowlege supported services for various real-world applications. Intelligent and intuitive interfaces of many smart objects, integrated within intelligent environment, interact with agents (humans, other devices, robots, etc.) in order to support their activities by information from the environment. On the other hand, intelligent environment is influenced by agents creating synergies between users and environment. The topic is focussed on identification of smart elements (devices, appliances) based on their behavior and interactions with other elements in the system. Analysis of the observed behavioral patterns of individual components will be the basis for creating predictive models (failures and deteriorations prediction, safety risks prediction, etc.). Artificial intelligence methods will be applied for the analysis of data, information and knowledge.

(Back)

Parallelization of computer simulation of fires
Supervisor: RNDr. Ján Glasa, CSc
Department: Department of Parallel Computational Methods and Algorithms
Advances in computer systems and information technologies have allowed the development of program systems capable to model complex phenomena related to fire. Such systems achieve a significant level of reliability and credibility. However, there is a lack of knowledge and experience related to parallelization of modelling fires in large spaces which is real motivation for this research. The research is focused on ways of parallelization which accelerate the calculation, but their impact on reliability and accuracy has not been sufficiently resolved yet. New knowledge applicable to solving practical problems of fire safety are expected. Calculations will be realized on high performance computer cluster at the Institute of Informatics, Slovak Academy of Sciences, Bratislava.

(Back)

Use of high performance computing for modeling of fire spread.
Supervisor: RNDr. Ján Glasa, CSc
Consultant: Ing. Lukáš Valášek, PhD.
Department: Department of Parallel Computational Methods and Algorithms
High performance computer systems allow to solve some tasks of fire safety in large spaces in which parallelization of simulation is necessary. The aim of the research is to investigate problems related to parallelization of fire simulation using paralel MPI model of the FDS system on computer clusters. The research will focus on efficiency and accuracy of simulation. Inspite of incresing demand for knowledge about the course of fires, only a limited number of papers dealing with this topic is available in the literature. New knowledge applicable to solving practical problems are expected. Calculations will be realized on high performance computer cluster at the Institute of Informatics, Slovak Academy of Sciences, Bratislava.

(Back)

Computer modeling of flows in road tunnel in fire conditions
Supervisor: RNDr. Ján Glasa, CSc
Consultant: Mgr. Peter Weisenpacher, PhD.
Department: Department of Parallel Computational Methods and Algorithms
Road tunnels belong to important parts of international transport systems. Therefore, fire safety issues in tunnels must be considered seriously. Research will focus on problems related to computer modeling of flows in road tunnel using the FDS program system which allows to model realistically and visualize flows caused by fire and simulate reaction of safety systems of the tunnel. Problems related to paralel realization of simulation on computer cluster will be investigated. Simulations will be realized on high performance computer cluster at the Institute of Informatics, Slovak Academy of Sciences, Bratislava.

(Back)

Distributed large data processing
Supervisor: Doc. Ing. L. Hluchý, CSc.
Department: Department of Parallel and Distributed Information Processing
Rapidly increasing volumes of diverse data from distributed sources create challenges for extracting valuable knowledge. Such applications can be considered modelling, simulation, pattern recognition, visualization, etc. in different areas as e.g. biomedicine, astrophysics, environmental sciences, aeronautics, automotive, energy, material sciences. Due to the size of the data, which are often referred to as large, extreme, it is necessary to design a methodology, robust methods and tools for extreme-scale analytics in synergy with distributed architectures for collecting and managing vast amounts of data such as Cloud Technologies and IoT. The dissertation project will be focused on the analysis, design of methodology, methods and algorithms for processing of large data for selected applications, which are currently solved at UI SAV. The research project will also include research and development of appropriate tools and services for distributed processing of methods and algorithms.

(Back)

Artificial Intelligence Methods in Cyber Security
Supervisor: Doc. Ing. L. Hluchý, CSc.
Department: Department of Parallel and Distributed Information Processing
Most current approaches to computer security focus on specific aspects of information and communication technology systems such as access control, cryptography, anonymization, virus protection, antivirus detection, intrusion detection, and anomaly detection. However, they lack an overall view of many aspects of cyber threats and do not pay due attention to one of the most important elements of cyber security: the human aspect. In addition, they often fail to address the dynamic nature of cyber attacks that are rapidly evolving and become more sophisticated by using new vulnerabilities and combining various attack channels (network, physical, human, etc.). To address these constraints and to increase our detection and response capabilities, we need a systematic and holistic approach to cyber security that takes into account technological and human factors. The dissertation project will focus on the design of methodology and methods for the analysis of anomalies and abnormalities using techniques of data acquisition and machine learning (data mining and processes mining) with the possibility of detection of hitherto unknown threats and vulnerabilities.

(Back)

Deep neural networks for applications in image processing and computer vision
Supervisor: Ing. Peter Malík, PhD.
Department: Department of Design and diagnostics of digital systems
Significant advances of artificial neural networks abilities in recent years have allowed their wider use in increasingly complex tasks. In a number of specific tasks, they surpassed the skills of a highly qualified human being – the expert, and more accurately evaluated presented data. Superhuman abilities were demonstrated in the image data classification of common objects, recognition and classification of human faces at various observation angles, as well as in classification of highly specific medical data from X-rays. The area of computer image processing has shown the greatest progress in reducing the error rate by using the specially designed and trained deep neural networks. Every year, new international open challenges focusing on a specific area of computer image processing and computer vision are announced. There is also growing interest from industry, and many of these challenges are directly supported by the private sector. One important factor in the success of artificial neural networks is a significant increase in the depth of modern neural network models in conjunction with a reduced number of parameters based on the convolutional principle and optimized architecture. Neural network architecture still offers wide space for improvement. There is also a lot of room for improvement in more complex tasks such as detection and instant segmentation where human abilities have not yet been overcome. Research will focus on these areas with the aim to develop new methods, algorithms, architectures to improve the capabilities of neural networks and their effective application in practice.

(Back)

Applied soft computing for complex solutions
Supervisor: Ing. Giang Nguyen, PhD.
Department: Department of Parallel and Distributed Information Processing
The thesis topic is focused on applied soft computing, which constructs computationally intelligent methods by combining edge technologies such as artificial neural networks, machine learning, deep learning, as well as optimization, fuzzy logic, and probabilistics to solve real problems. Various methods used in soft computing are neither independent of nor compete with one another, but rather, they work in a cooperative and complementary way. Soft computing aims to tolerance of imprecision, uncertainty, partial truth and approximation to achieve effectiveness and low solution cost. When today data has large-scale potencial with Volume, Velocity, Variety and Veracity characteristics, the wider collaboration between intelligent soft computing methods, large-scale data preprocessing and high-performance backgrounds is practical to face challenges in many areas. All of these advanced technologies do not have to be always coupled together, but the alliance is essential for complex solutions.

(Back)

Sentiment analysis in text
Supervisor: Ing. Giang Nguyen, PhD.
Consultant: Ing. Štefan Dlugolinský, PhD.
Department: Department of Parallel and Distributed Information Processing
The thesis topic is focused on methods for sentiment analysis in text resources. It is an opinion mining, most often from product reviews, microblogs or from news articles in order to get marketing information, or overview of the opinion spectrum respectively. The thesis will continue from the state-of-the-art, it will design and validate appropriate methods applicable to Slovak language. However, the topic is not limited only to the Slovak language. In addition, the student may also explore methods suitable for other languages. The topic is closely related to areas such as recommendation, user-modelling, etc. offering a variety of approaches to be taken such as deep learning, natural language processing or parallel and distributed data processing such as Hadoop or Spark.

(Back)

FURHAT Robotics’s Robotic Head as an Expressive Voice Assistant input/output device in Slovak
Supervisor: Ing. Milan Rusko, PhD.
Department: Department of Speech analysis and synthesis
The FURHAT Robotic Head uses animated image on a human-face-like screen to create the illusion of a talking human head. It provides the ability to use expressive facial expressions as well asi t is equipped with input and output audio devices. The student will give an overview of the state of the art in expressive speech research. He will define the emotions that an assistant should be able to represent and, using the FURHAT head, as well as speech speech synthesizer and recognizer developed on UI SAV, he will create an I/O for the expressive speech assistant in Slovak.

(Back)

Automatic information retrieval using the natural language processing
Supervisor: Ing. Milan Rusko, PhD.
Department: Department of Speech analysis and synthesis
The theme focuses on the area of text, document and web processing. The aim is to analyze, create and verify methods and algorithms to solve information extraction tasks, recognition of names and other entities as well as processing of available resources from the web and their use in methods of statistical processing of natural language. It is possible to specialize in the processing of Slovak or English language or multilingual approaches.

(Back)

Deep neural networks in end-to-end speech recognition
Supervisor: Ing. Milan Rusko, PhD.
Department: Department of Speech analysis and synthesis
The introduction of Deep Neural Networks into speech recognition systems has significantly improved the reliability of the recognizers and their resistance to inter-speaker and intra-speaker variability, and various transmission channel characteristics. The PhD student will give an overview of this new approach and its use in the speech recognition, including end-to-end approaches. Using existing speech and text databases developed at UI SAV, he will design and implement the end-to-end speech recognizer in Slovak and compare it to the previous approaches.

(Back)

Automatic measurement of stress from voice
Supervisor: Ing. Milan Rusko, PhD.
Department: Department of Speech analysis and synthesis
The acoustic characteristics of the speech signal reflect, among other information, the emotional state of the speaker. The role of the doctoral student is to design a recording method/scenario to record an expressive speech database containing increased stress, for example, using computer game or virtual reality. The collected database will be used to train a stress-level detection system based on machine learning methods. The acoustic cues of stress in the speech signal will be analysed.

(Back)

Application development, deployment and orchestration in Cloud computing
Supervisor: Ing. Dinh Viet Tran, PhD.
Department: Department of Parallel and Distributed Information Processing
With the advance of cloud technologies, the trend of developing and deploying services/applications in cloud environment also has appeared. There are economic as well as technological reasons why an application should be developed and deployed on the cloud. On the economic side, cloud computing can provide significant cost savings due to the increased utilization resulting from the pooling of resources (often virtualized). Furthermore, cloud computing enables rapid delivery of IT services, which increases business efficiency. This is the reason, Cloud Computing attracts large enterprises/companies to build and provide outside Cloud services in order to make a profit. However, as Cloud applications become more and more complex and many types of cloud computing may be in use simultaneously, it is challenging to create, deploy and manage these applications across different Cloud infrastructures. The aim of the thesis is to propose an approach, method and tools for solving the problem of application development, deployment and orchestration in multi-cloud environment.

(Back)