Wednesday, June 26, 2024 | 10:00 am (CET) | Room: HS 11 | | Alpen-Adria-Universität Klagenfurt

Dr. Farzad Tashtarian | Department of Information Technology (ATHENA Christian Doppler (CD) Laboratory)

Abstract: In today’s digital landscape, video streaming has become a cornerstone of online content delivery, with applications spanning entertainment, education, teleconferencing, and live broadcasting. The importance of optimizing video streaming cannot be overstated, as Quality of Experience (QoE) directly impacts user engagement and satisfaction. However, the video streaming pipeline faces significant challenges, including maintaining high QoE, minimizing latency, and managing costs effectively. This talk addresses these critical issues by exploring holistic strategies that enhance the video streaming experience from source to destination. We will delve into the complexities of content provisioning, content delivery, and content consumption, emphasizing the necessity of an end-to-end approach. Additionally, by considering the entire streaming pipeline, we will discuss novel algorithms designed for optimizing end-to-end video streaming and demonstrate how these approaches can more effectively tackle the primary challenges of QoE, end-to-end latency, and video streaming costs.

Bio: Farzad Tashtarian is a post-doctoral researcher in the ATHENA project at the Institute of Information Technology (ITEC), Alpen-Adria-Universität Klagenfurt (AAU). Prior to joining ATHENA, he served as an Assistant Professor at Azad University of Mashhad, Iran. He earned his Ph.D. in Computer Engineering from Ferdowsi University of Mashhad, Iran. As a researcher, he has co-authored more than 60 papers published in prestigious journals and conferences. He is a work-package leader and scientific coordinator in the ATHENA Projects at Alpen-Adria-Universität Klagenfurt. His current research interests include Multimedia Networks, Immersive Media, Generative AI for Multimedia Streaming, software-defined networking, mathematical modeling, and distributed optimization. For more information visit: https://tashtarian.net/

 

Tuesday, June 25, 2024 | 09:00 am (CET) | Room: HS 10 | Alpen-Adria Universität Klagenfurt

Prof. Tamer Khatib | An-Najah National University

Abstract: In order to assure sustainable development, decentralized green management is required through local government units. Such a process required competences that imply behavioral, structural and strategic aspects. Join us to learn more on how to green governments for localized sustainable development.

Bio: Tamer is a photovoltaic power systems professional. He holds a B.Sc. degree in electrical power systems from An-Najah National University, Palestine as well as a M.Sc. degree and a Ph.D degree in photovoltaic power systems from National University of Malaysia, Malaysia. In addition he holds Habilitation (the highest academic degree in German speaking countries) in Renewable and sustainable energy from Alpen Adria Universitat, Klagenfurt, Austria. Currently he is a Full Professor of renewable energy at An-Najah National University. He is also the director of An-Najah Company for Consultancy and Technical Studies. So far, he has 2 patents, 4 books and 140 research articles, while his current h-index is 42. He has supervised 5 Ph.D researches, 19 master researches and 65 bachelor researches. He is a senior member of IEEE, IEEE Power and Energy Society, The International Solar Energy society, Jordanian Engineers Association, and International Association of Engineers.

His research interests mainly fall in the scope of photovoltaic systems and solar energy fundamentals. These interests include PV systems design and optimization, modeling and control of PV systems, hybrid PV/Wind systems, hybrid PV/diesel systems, Grid connected PV systems, sun trackers, MPPT technology, inverters in PV system, solar chargers, batteries and charge controllers, solar energy fundamentals, solar energy prediction, AI applications for solar energy and PV systems, wind power systems, wind chargers, wind energy modeling and prediction.

Find the SLIDES on on our slideshare account: https://www.slideshare.net/slideshow/greening-local-government-units-current-status-and-required-competences/269872823

 

Thursday, June 13, 2024 | 03:00 pm (CET) | Room: S.2.42 | Alpen-Adria Universität Klagenfurt

Dr. Daniel Thilo Schroeder | Researcher at SINTEF

Abstract: In an era where efficient processing of complex graph structures is increasingly vital across various domains, such as social networks and biological system modeling, there arises a need for robust computational architectures. This Kolloquium introduces a novel approach that incorporates a higher-order pipeline architecture, aiming to enhance graph data processing. Integrating functional programming with object-oriented principles facilitates complex data processing through an intuitive, modular system. Central to this system is the notion of treating computation units as first-class entities, which promotes modular and type-safe environments. Additionally, the system includes a higher-order traversal abstraction to support flexible data manipulation strategies, a directed data-transfer protocol for efficient data flow management, and an operator model that enhances the robust lifecycle management of operations. The adaptability and performance of this system are further augmented by its capability to incorporate various graph computation models, such as vertex-centric and edge-centric processing. This approach contributes to scalable solutions that meet diverse computational needs without overcomplicating the user interface by offering a structured yet adaptable method for graph data handling.

Bio: Daniel Thilo Schroeder is a Research Scientist at SINTEF’s Smart Data group, where he develops high-performance, scalable platforms for processing and analyzing extreme data using massive graph representations, facilitated by serverless computing. He is actively involved in the enRichMyData project, focusing on creating FAIR-compliant datasets essential for Big Data and AI applications, and the GraphMassivizer Project, which seeks to revolutionize data processing through sustainable, advanced graph technologies. Previously, during his postdoctoral research at Simula’s Department for High-Performance Computing, he developed new computational frameworks for Graph Neural Networks to enhance deep learning for unstructured data. Daniel earned his PhD by focusing on the early detection and analysis of online misinformation, aiming to develop strategies for its mitigation. Contact Daniel at daniel.t.schroeder@sintef.no or visit https://danielthiloschroeder.org/

 

 

Thursday, April 25, 2024 | 03:30 pm (CET) | Room: S.2.69 | Alpen-Adria Universität Klagenfurt

Alexandru Iosup | University Research Chair and Full Professor at the Vrije University of Amsterdam

Abstract: Our society has turned digital: From science to business, from online shopping to online gaming, from education to government, digital applications depend every moment on Information and Communication Technology (ICT) infrastructure. To pilot this infrastructure and navigate through it, we need detailed information and decision-making support, provided on-time, cheaply, and reliably. (We briefly argue non-data-driven approaches are currently unable to cope with real conditions.) In this talk, we argue that Operational Data Analytics (ODA) can provide these capabilities, with advances not only in monitoring and observability, but also in data sourcing and ontology mapping, data cleaning and filtering, data characterization and modeling, and process management. Furthermore, we argue a digital twin, capable of simulating both the current conditions and making long-term predictions, should be integrated with the ODA system – providing an advanced form of A – to help with resource management and scheduling, pipeline optimization, energy awareness, general system tuning, capacity planning, etc. We present a reference architecture for ODA, a partial analysis of the state of the art, and experience with data collection and its use in the digital twinning of datacenters. This work aligns with and benefits from collaboration with the SPEC RG Cloud Group and, among others, the EU Graph-Massivizer project, and the OffSense and 6G Future Network Services projects in the Netherlands.

Bio: Dr.ir. Alexandru Iosup is a full professor at Vrije Universiteit Amsterdam (VU), and the chair of the VU research group Massivizing Computer Systems and the VU Computer Science Honours Programme. He is also co-chair of the national IPN SIG on Future Computer Systems and Networking and elected chair of the international SPEC-RG Cloud Group. His fascination for massivizing distributed systems and ecosystems across the digital computing continuum, with applications in cloud computing, big data and graphs, scientific and business-critical computing, and the metaverse, has received prestigious recognition, including membership in the (Young) Royal Academy of Arts and Sciences of the Netherlands, the Netherlands ICT Researcher of the Year award, and a Ph.D. from TU Delft. These accompany over 200 publications, and numerous open science FAIR artifacts such as software and datasets and archives, with high scientific and technical impact. Further, dr. Iosup’s leadership and innovation in education led to various awards, including the prestigious Netherlands Higher Education Teacher of the Year. He has also received a knighthood for cultural and scientific merits. Contact Alexandru at A.Iosup@vu.nl or visit http://atlarge.science/aiosup

 

Thursday, April 25, 2024 | 08:30 am (CET) | Room: S.2.37 | Alpen-Adria Universität Klagenfurt

Ana Lucia Varbanescu | Assoc. Prof. Dr. ir. at the University of Twente and University of Amsterdam

Abstract: “Computation” has become a massive part of our daily lives: in science, a lot of experiments and analysis rely on massive computation, in AI we use vast resources to train and use massive models, and in engineering we use complex simulations and digital twins to increase efficiency and productivity. Under the assumption that computation is cheap, and time-to-result is the only relevant metric, we often use significant computational resources at low efficiency. In this talk, I argue this approach is an unacceptable waste of computing resources, and demonstrate we can do better! By means of a couple of case-studies, I will show how performance engineering can be used for zero-waste computing, proving how efficiency and time-to-result can be happily married. I will further propose a co-design methodology that leverages such performance engineering methods to enable the selection of algorithms _and_ their effective deployment on suitable infrastructure. The approach relies on design-space exploration, driven by efficient search methods and compositional performance models. I will conclude by reflecting on the next steps and open questions that need answers to make this co-design approach feasible and applicable for more applications and systems.

Bio: Ana Lucia Varbanescu holds a BSc and MSc degree from POLITEHNICA University in Bucharest, Romania. She obtained her PhD from TU Delft, The Netherlands, and continued to work as a Postdoc researcher in The Netherlands, at TU Delft and VU University in Amsterdam. She is a MacGillavry fellow at University of Amsterdam, where she was tenured in 2018 as Associate Professor. Since 2022, she is Professor at University of Twente, The Netherlands. She has been a visiting researcher at IBM TJ Watson (2006, 2007), Barcelona Supercomputing Center (2007), NVIDIA (2009), and Imperial College of London (2013).  She has received several national grants (including a personal Veni grant) and she is a co-PI for the Graph-Massivizer EU project. Ana’s research stems from HPC, and investigates the use of heterogeneous systems for high-performance applications, with a special focus on performance and energy efficiency modeling for both scientific and data-intensive applications. Her latest research focuses on zero-waste computing and model-based systems co-design.

 

Friday, March 15, 2024 | 10:00 am (CET) | Room: S.2.69 | Universität Klagenfurt Matteo Zavatteri | Assistant Professor  in computer science at the Department of Mathematics of University of Padua, Italy Abstract: A recent direction in Business Process Management studied methodologies to control the execution of Business Processes under several sources of uncertainty in order to always get to the end by satisfying all constraints. Current approaches encode business processes into temporal constraint networks or timed game automata in order to exploit their related strategy synthesis algorithms. However, the proposed encodings can only synthesize single-strategies and fail to handle loops. To overcome these limits I will discuss a recent approach based on supervisory control. The approach considers structured business processes with resources, parallel and mutually exclusive branches, loops, and uncertainty. I will discuss an encoding into finite state automata and prove that their concurrent behavior models exactly all possible executions of the process. After that, I will introduce tentative commitment constraints as a new class of constraints restricting the executions of a process. Finally, I will discuss a tree decomposition of the process that plays a central role in modular supervisory control. Bio: Matteo Zavatteri is assistant professor in computer science at the Department of Mathematics of University of Padua, Italy. He received a master degree in engineering and computer science and a Ph.D. in computer science from University of Verona (ITALY). His research interests are in the fields of formal methods, discrete event systems, supervisory control, temporal reasoning, artificial intelligence, and business process management. Currently he works in the iNEST (Interconnected Nord-Est Innovation Ecosystem) project, which is a project supported with 110M€ by the National Recovery and Resilience Plan, funded by the European Union in the NextGenerationEU plan. In such a project he develops neuro-symbolic approaches for digital twins.

 

Wednesday, February 28, 2024 | 15:30 pm (CET) | Room: S.2.42 | Alpen-Adria Universität Klagenfurt

Amr Rizk | Professor at the Department for Computer Science at the University of Duisburg-Essen

Abstract: This talk describes a vision of processing-centric networks that unify In-Network computation and programmable communication. This convergence of computation and communication enables a high degree of flexibility in the design and operation of distributed applications that run „on top“ or „within“ these networks. We specifically consider some of these applications that must adhere to given performance requirements, e.g., on timing, with a main focus on Multimedia Systems. Major challenges to the design and operation of processing-centric networks lie in the coupling of the programmable control of networks with application and network models that are fed with real-time feedback data. The aim of such coupling is to provide flexibility in combination with provable performance guarantees.

Bio: Amr Rizk received the doctoral degree (Dr.-Ing.) from the Leibniz Universität Hannover, Germany, in 2013. After that he held postdoctoral positions at University of Warwick, UMass Amherst and the TU Darmstadt, Germany. From 2019 to 2021 he was an assistant professor at Ulm University, Germany.

Since 2021 he is a professor at the department for computer science at the University of Duisburg-Essen, Germany.  He was a recipient of the Excellence in DASH Award at ACM MMSys 2016 and the Best Paper Award at ACM Middleware 2017 and ACM MMSys 2023. He is interested in performance evaluation of communication networks, stochastic models, and their applications to distributed systems.

 

Tuesday, February 13, 2024 | 10:00 am (CET) | Room: S.1.42 | Alpen-Adria Universität Klagenfurt

Jean-Charles Régin| Professor at the Université Côte d’Azur and Chair for Decision Intelligence at the Interdisciplinary Institutes of Artificial Intelligence (3IA)

Abstract: This talk introduces a new approach to generate strongly constrained texts. We consider standardized sentence generation for the typical application of vision screening. To solve this problem, we formalize it as a discrete combinatorial optimization problem on words and show how constraint programming and multivalued decision diagrams (MDD), a well-known data structure to deal with constraints can be used to solve it. We show how part of the language is kept thanks to n-grams. Once the sentences are obtained, we apply a language model (LLM: GPT-3) to keep the best ones. We detail this for English and also for French where the agreement and conjugation rules are known to be more complex. We will also discuss about some possible improvements for a better integration of LLM into constraint programming.

(This presentation, does not require any knowledge in LLM or in constraint programming.)

Bio: JC Régin is an internationally recognized expert of Constraint Programming (CP). Innovation and ground-breaking research are a constancy in his career and his contributions are recognized internationally in the academic and non-academic worlds. He is one of the inventors of the global constraints in CP (i.e. algorithms to quickly eliminate incoherent values). His paper describing the All different global constraint is highly cited (> 1000 cit.). This constraint is now implemented in almost all CP solvers and routinely used in many applications by companies such as IBM, Oracle, Microsoft, Google or SAS. The cited article, published in 1994, received the „Classical Paper Award“ in 2013 from the American Association on Artificial Intelligence: „For ground-breaking contributions to constraint programming via the development of one of the first propagators for global constraints.“

In 2013, he received the Association for Constraint Programming Research Excellence Award „in recognition of a program of seminal and outstanding scientific contributions to both the theory and practice of constraint programming“.

Thursday, February 8, 2024 | 09:00 am (CET) | Room: S.2.37 | | Alpen-Adria-Universität Klagenfurt

Dr. Hadi Amirpourazarian BSc.MSc. | Department of Information Technology (ATHENA Christian Doppler (CD) Laboratory)

Abstract: The widespread adoption of video streaming applications has experienced a continual upsurge, necessitating advancements in underlying technologies to ensure a seamless user experience. Central to the realm of video streaming is video coding, a key element in the efficient transmission of multimedia content.

In the context of HTTP Adaptive Streaming (HAS), videos are encoded at multiple bitrate-resolution pairs, collectively forming what is known as the bitrate ladder. This approach allows users to adapt to varying network conditions and select the most appropriate bitrate-resolution pair for a given content, thereby enhancing the overall Quality of Experience (QoE). This research extends its focus beyond (i) enhancing individual bitrate-resolution pairs to also include (ii) optimizing the construction of the entire bitrate ladder. For the former, innovative solutions utilizing content-aware deep neural networks are proposed. The primary objective is to elevate the quality of single bitrate-resolution pairs through the application of advanced deep learning techniques tailored to the unique characteristics of the content.

In pursuit of the latter objective, solutions are sought to determine the number of bitrate-resolution pairs in the bitrate ladder and their corresponding encoding parameters, such as optimal bitrate and resolution. (i) Network-assisted or (ii) content-aware approaches may be employed to find the optimal number of bitrate-resolution pairs, while content-aware solutions can assist in determining optimal resolutions, frame rates, and other parameters for the selected bitrates. Attention is also given to online bitrate ladder constructions, addressing the latency challenges inherent in such scenarios.

Recognizing that video coding is incomplete without proper assessment of quality, efforts are directed towards integrating video quality and coding. Real-time quality metrics are identified, and key parameters influencing Quality of Experience (QoE) are considered. Additionally, solutions are explored in video coding and transcoding to reduce encoding time/costs and energy consumption while minimizing quality degradation. This holistic approach aims to contribute comprehensively to the enhancement of video streaming technologies.

Bio: Hadi Amirpour is a postdoctoral research fellow at the Christian Doppler (CD) Laboratory ATHENA, based at the University of Klagenfurt. He earned his Ph.D. in computer science from the University of Klagenfurt in 2022. He holds two B.Sc. degrees in Electrical and Biomedical Engineering, as well as an M.Sc. degree in Electrical Engineering from K. N. Toosi University of Technology. Previously, he was part of the EmergIMG project, a Portuguese consortium focusing on emerging imaging technologies, funded by the Portuguese funding agency and H2020. His research interests encompass video streaming, image and video compression, quality of experience, emerging 3D imaging technology, and medical image analysis. Hadi has actively participated in standardization committees such as JPEG Pleno and MPEG. Currently, he serves as the co-chair of Qualinet TF7 since 2021.

Additionally, he has authored more than 80 publications and patents, including contributions to high-prestige journals such as IEEE  COMST, IEEE TIP, IEEE TMM, and IEEE TCSVT. He has also contributed to the organization of special sessions, workshops, etc., at international conferences, including ACM Multimedia 2022, IEEE EUVIP 2022, ACM MobiSys 2022, IEEE ICME 2023. Furthermore, he has contributed to the academic community by giving two tutorials at IEEE ICME 2023 and IEEE VCIP 2023.

 

Friday, January 19, 2024 | 12:15 – 13:30 pm | Room: Z.1.29 | AAU Klagenfurt

Special Guest & Input Klaus Diepold (TU München)

Welcome & Chair Martina Merz, Christian Timmerer

Klaus Diepold is a professor at the Technical University of Munich and works in the School of Computation, Information, and Technology; he holds the Chair in Data Processing since 2002. He conducted research in the field of multimedia signal processing with a focus on the design of fast algorithms; today, he concentrates on machine learning, including the development of moral machines and the modeling and simulation of emotions for cognitive systems. At the Center for Digital Technology and Management (CDTM), he is also involved in entrepreneurship education for students from all disciplines. Klaus Diepold will give a lecture on the topic „How to Connect, Educate and Empower the Innovators of Tomorrow“.

What Are the Klagenfurt Fireside Chats?

The fireside chats start with the question of how we can transfer knowledge and technology from the scientific communities into the economy and society. Special guests with experience in knowledge and technology transfer give an introductory speech. Afterward, scholars exchange ideas informally and openly in a small circle. The talks are intended to explore the potential for activities and cooperation in knowledge and technology transfer and generate surprising insights.

We are very much looking forward to your participation!
Martina Merz, vice-rector for research