About

SerbianAI is an initiative started by Prof. Predrag Janicic and Pavle Subotic after a successful mini workshop in 2019. We are a informal networking group that meets to organise workshops, meetings and promote collaboration between AI researchers in Serbian and abroad. Moreover, we would like to engage local industry to support local academics. We strongly feel the best way to reduce the brain drain is to strengthen our local scientific institutions so that our brightest minds can build their research careers in Serbia.

СербианАИ је иницијатива коју су покренули проф. Предраг Јаничић и Павле Суботић након успешне мини радионице 2019. године. Ми смо неформална група за умрежавање која се састаје ради организовања радионица, састанака и промоције сарадње између истраживача АИ у Србији и иностранству. Штавише, желели бисмо да ангажујемо локалну индустрију да подржи локалне академике. Снажно осећамо да је најбољи начин за смањење одлива мозгова јачање наших локалних научних институција како би наши најбистрији умови могли да граде своју истраживачку каријеру у Србији.

Meeting 2020

Video na YouTube kanalu

Program / Програм:

10:00 Otvaranje - Predrag Janicic i Pavle Subotic

10:10 Dijaspora -- Logika i formalne metode
(predsedava Pavle Subotic)

Alen Arslanagic, U Groningen, Holland
Stefan Jaksic, AIT Vienna, Austria
Dejan Jovanovic, SRI, USA
Marijana Lazic, TU Munich, Germany
Ranko Lazic, U Warwick, UK
Petar Maksimovic, Imperial College London, UK
Marija Selakovic i LazarStricevic, Huawei, Dresden, Germany
Petar Vukmirovic, Vrije U, Amsterdam, Holland
Djordje Zikelic, IST, Austria

10:30 Dijaspora -- Obrada prirodnog jezika
(predsedava Pavle Subotic)

Milica Gasic, U Dusselforf, Germany
Vlado Keselj, Dalhousie U, Canada
Nikola Milosevic, Bayer, Germany
Jelena Mitrovic, U Passau, Germany
Maja Popovic, Dublin City U, Ireland
Irena Spasic, U Cardiff, UK
Goran Nenadic, U Manchester, UK

10:45 Dijaspora: Masinsko ucenje
(predsedava Pavle Subotic)

Djordje Gligorijevic i Jelena Gligorijevic, Yahoo Research, Sunnyvale, USA
Vladimir Mitrovic, Amazon, USA
Andrija Tomovic, Novartis, Switzerland
Petar Velickovic, DeepMind, UK

10:55 Pauza

11:00 Akademske institucije u Srbiji
(predsedava Predrag Janicic)

ETF-Bg, Drazen Draskovic
FON-Bg, Vladan Devedzic
FLF-RGF-JERTEH-Bg, Ranka Stankovic
FTNCacak-Kg, Vladimir Mladenovic
FTN-NS, Jelena Slivka
Inf-PMF-NS, Milos Radovanovic
Mas(CAM)-Ni, Melanija Mitrovic
MI-SANU, Zoran Ognjanovic
MatF-Bg, Filip Maric
DU-NoviPazar, Ulfeta Marovac
PMF-Kg, Bodan Stojanovic
PMF-Ni, Miroslav Ciric
PMF-Pr, Stefan Panic
U Singidunum, Milos Stankovic
MDCS, Andreja Ilic
RCC, Damjan Krstajic
Mat-PMF-NS, Andreja Tepavcevic

11:30 Pauza

11:45 Panel diskusija

Slides / Слајдови

Participants / Учесници:

Logic and Formal Reasoning / Логика и формалне методе:

Petar Maksimović	Imperial College London
Ranko Lazic	University of Warwick
Marija Selakovic	Huawei Research
Petar Vukmirovic	Vrije Universiteit Amsterdam
Predrag Janičić	MatF, University of Belgrade
Djordje Zikelic	IST Austria
Pavle Subotic	psubotic.github.io
Vesna Marinkovic	MatF, University of Belgrade
Tatjana Lutovac	ETF, University of Belgrade
Zoran Ognjanović	Mathematical institute SANU
Andreja Tepavčević	Mathematical institute SANU & PMF University of Novi Sad
Miroslav Ćirić	PMF, University of Niš
Milena Vujosevic Janicic	MatF, University of Belgrade
Filip Marić	MatF, University of Belgrade
Vojin Jovanovic	Oracle Labs
Marijana Lazic	TU Munich
Silvia Gilezan	FTN University Novi Sad, MISANU
Sana Stojanovic Đurđević	MatF, University of Belgrade
Danijela Simić	MatF, University of Belgrade
Stefan Stanimirović	PMF, University of Niš
Simona Kašterović	FTN, University of Novi Sad
Branko Malešević	Univerzitet u Beogradu - Elektrotehnički fakultet (ETF)
Tamara Stefanović	FTN, University of Novi Sad
Alen Arslanagic	Uni. Groningen
Lazar Stričević	Huawei Dresden Research Center
Stefan Jakšić	Austrian Institute of Technology
Tatjana Stojanović	PMF, University of Kragujevac

NLP / Обрада природног језика:

Anđelka Zečević	MatF, University of Belgrede
Lazar Davidović	University of Belgrade
Vladisav Jelisavcic	Mathematical Institute SANU
Branko Arsić	PMF, University of Kragujevac
Nikola Milosevic
Danka Jokic	University of Belgrade
Marko Putniković	FON, University of Belgrade
Ranka Stanković	RGF, University of Belgrade
Branislava Šandrih	Fil, University of Belgrade
Jelena Mitrović	University of Passau
Irena Spasic	Cardiff University
Aleksandar Kovacevic	FTN, University of Novi Sad
Goran Nenadic	University of Manchester
Maja Popović	Dublin City University
Aldina Avdić	State University of Novi Pazar
Adela Ljajić	State University of Novi Pazar
Ulfeta Marovac	State University of Novi Pazar
Vlado Keselj	Dalhousie University, Faculty of Computer Science
Miloš Stanojević	University of Edinburgh

ML / Машинско учење:

Irfan Fetahovic	Drzavni univerzitet u Novom Pazaru
Dragana Radojicic	Mathematical institute SANU
Mladen Nikolic	MatF, University of Belgrade
Vladimir Urošević	MatF, University of Belgrade
Gorana Gojić	Faculty of Technical Sciences, University of Novi Sad
Predrag Tadić	University of Belgrade - School of Electrical Engineering (ETF)
Igor Ilic	Microsoft
Damjan Dakić	Microsoft
Andreja Ilic	Microsoft
Miloš Radovanović	PMF, University of Novi Sad
Djordje Gligorijevic	Yahoo Research
Jelena Gligorijevic	Yahoo Research
Radovan Turović	Faculty of Technical Sciences, University of Novi Sad
Petar Veličković	DeepMind
Milos Stankovic	University Singidunum. Belgrade

General AI / Општа вештачка интелигенција:

Vladan Devedzic	FON, University of Belgrade
Melanija Mitrović	Faculty of Mechanical Engineering, Niš
Bogdan Djordjević	Mathematical Institute SANU
Tatjana Davidović	Mathematical Institute SANU
Vladimir Mitrovic	Amazon, AWS AI (Elastic Inference)
Vladimir Mladenovic	FTN, University of Kragujevac
Jelena Slivka	Faculty of Technical Sciences, University of Novi Sad
Tatjana Jakšić Krüger	Mathematical institute SANU
Velimir Ilić	Mathematical institute SANU
Stefan Panic	PMF, University of Pristina
Negovan Stamenkovic	PMF, University of Pristina
Dušan Gajić	Faculty of Technical Sciences, University of Novi Sad
Irfan Fetahović	State University of Novi Pazar
Ivana Štajner-Papuga	Univerzitet u Novom Sadu
Dražen Drašković	Univerzitet u Beogradu - Elektrotehnički fakultet (ETF)
Boban Stojanović	PMF, University of Kragujevac
Ana Kaplarević-Mališić	PMF, University of Kragujevac
Miloš Ivanović	PMF, University of Kragujevac
Višnja Simić	PMF, University of Kragujevac
Andrija Tomovic	Novartis, Switzerland
Nina Radojičić Matić	MatF, University of Belgrade
Branimir Seselja	PMF, University of Novi Sad
Miroslav Marić	MatF, University of Belgrade
Jelena Ignjatović	PMF, University of Niš
Branislav Kisačanin	Nvidia
Damjan Krstajic	Research Centre for Cheminformatics, Serbia

Seminars 2021

МАТЕМАТИЧКИ ИНСТИТУТ САНУ, Београд
СЕМИНАР ИЗ ВЕШТАЧКЕ ИНТЕЛИГЕНЦИЈЕ
ПРОГРАМ ЗА АПРИЛ 2021

1. 7. април 2021. 19-20h
   др Предраг Јаничић
   редовни професор, Математички факултет, Универзитет у Београду
   
   Наслов: Аутоматско резоновање и примери система за резоновање у исказној логици, у логици првог реда и у геометрији
   
   Апстракт: У предавању ће бити дат кратак приказ области аутоматског резоновања, посебно централне подобласти – аутоматског доказивања теорема, као и неких области примена. Биће укратко приказано и неколико предавачевих системa који користе аутоматско резоновање у исказној логици, у логици првог реда и у геометрији и биће описане неке њихове конкретне примене.
   
   
2. 14.април 2021. 19-20h
   др Павле Суботић
   Senior Research Engineer at Azure Data Labs
   
   Title: Debugging Large Scale Datalog with Proof Annotations
   
   Abstract: Logic programming languages such as Datalog have become popular as Domain Specific Languages (DSLs) for solving large-scale, real-world problems, in particular, static program analysis, graph databases and network analysis. The logic specifications that model analysis problems process millions of tuples of data and contain hundreds of highly recursive rules. As a result, they are notoriously difficult to debug. While the database community has proposed several data provenance techniques that address the Declarative Debugging Challenge for Databases, in the cases of analysis problems, these state-of-the-art techniques do not scale. In this talk, I introduce a novel bottom up Datalog evaluation strategy for debugging: Our provenance evaluation strategy relies on a new provenance lattice that includes proof annotations and a new fixed-point semantics for semi-naïve evaluation. A debugging query mechanism allows arbitrary provenance queries, constructing partial proof trees of tuples with minimal height. We integrate our technique into Soufflé, a Datalog engine that synthesizes C++ code, and achieve high performance by using specialized parallel data structures. Experiments are conducted with DOOP/DaCapo, producing proof annotations for tens of millions of output tuples. We show that our method has a runtime overhead of 1.31× on average while being more flexible than existing state-of-the-art techniques. This is joint work with David Zhao and Prof. Bernhard Scholz from the University of Sydney. A version of this talk was presented at POPL this year (2021).
   
   
3. 21.април 2021. 19-20h
   др Бранислав Кисачанин
   Nvidia + Institut br.ai.ns
   
   Наслов: Рачунарске архитектуре за AI/ML: Када је која најбоља?
   
   Апстракт: У овом предавању ћемо се позабавити питањем са којим се суочава већина AI/ML истраживача, о томе која је рачунарска архитектура оптимална за проблем који решавају. Да ли је то GPU, CPU, FPGA, DSP, ASIC или нешто сасвим ново? Можемо ли AI да применимо на паметном телефону? Питања има много а правих одговора мало чак и на Гуглу.
   
   
4. 28.април 2021. 19-20h
   др Петар Величковић
   Senior Research Scientist at DeepMind
   
   Title: Geometric Deep Learning: Grids, Graphs, Groups, Geodesics and Gauges
   
   Abstract: The last decade has witnessed an experimental revolution in data science and machine learning, epitomised by deep learning methods. Indeed, many high-dimensional learning tasks previously thought to be beyond reach –such as computer vision, playing Go, or protein folding – are in fact feasible with appropriate computational scale. Remarkably, the essence of deep learning is built from two simple algorithmic principles: first, the notion of representation or feature learning, whereby adapted, often hierarchical, features capture the appropriate notion of regularity for each task, and second, learning by local gradient-descent type methods, typically implemented as backpropagation. While learning generic functions in high dimensions is a cursed estimation problem, most tasks of interest are not generic, and come with essential pre-defined regularities arising from the underlying low-dimensionality and structure of the physical world. This talk is concerned with exposing these regularities through unified geometric principles that can be applied throughout a wide spectrum of applications. Such a ‘geometric unification’ endeavour in the spirit of Felix Klein's Erlangen Program serves a dual purpose: on one hand, it provides a common mathematical framework to study the most successful neural network architectures, such as CNNs, RNNs, GNNs, and Transformers. On the other hand, it gives a constructive procedure to incorporate prior physical knowledge into neural architectures and provide principled way to build future architectures yet to be invented.