#Accepted Talks

Automated Harmful Content Detection Using Grammar-Focused Representations of Text Data (slides)
Daria Stetsenko (NASK PIB), Inez Okulska (NASK PIB)

Building more helpful and safe dialogue agents via targeted human judgements
Maja Trębacz (DeepMind) on behalf of all the authors: Amelia Glaese*, Nat McAleese*, Maja Trębacz*, John Aslanides*, Vlad Firoiu, Timo Ewalds, Maribeth Rauh, Laura Weidinger, Martin Chadwick, Phoebe Thacker, Lucy Campbell-Gillingham, Jonathan Uesato, Po-Sen Huang, Ramona Comanescu, Fan Yang, Abigail See, Sumanth Dathathri, Rory Greig, Charlie Chen, Doug Fritz, Jaume Sanchez Elias, Richard Green, Soňa Mokrá, Nicholas Fernando, Boxi Wu, Rachel Foley, Susannah Young, Iason Gabriel, William Isaac, John Mellor, Demis Hassabis, Koray Kavukcuoglu, Lisa Anne Hendricks, Geoffrey Irving (DeepMind), * - Equal contributions

Consolidated learning – new approach to domain-specific strategy of hyperparameter optimization
Katarzyna Woźnica (MI2DataLab; Warsaw University of Technology)

A Fictitious Play Approach for Multiplayer Computer Games (slides)
Jacek Cyranka (University of Warsaw), Jarek Kochanowicz, Witold Szejgis

Fast and Precise: Adjusting Planning Horizon with Adaptive Subgoal Search (slides)
Michał Zawalski (University of Warsaw), Michał Tyrolski (University of Warsaw), Konrad Czechowski (University of Warsaw), Damian Stachura (Jagiellonian University), Piotr Piękos (University of Warsaw), Tomasz Odrzygóźdź (IDEAS NCBR), Yuhuai Wu (Google Research; Stanford University), Łukasz Kuciński (Polish Academy of Sciences), Piotr Miłoś (Polish Academy of Sciences; deepsense.ai)

Fine-Grained Conditional Computation in Transformers (slides)
Sebastian Jaszczur (IDEAS NCBR ; University of Warsaw)

From data acquisition to ML model for predicting outcomes of chemical reactions (slides)
Michał Sadowski (Molecule.one), Mikołaj Sacha (Molecule.one)

How to learn classifier chains using positive-unlabelled multi-label data? (slides)
Paweł Teisseyre (Polish Academy of Sciences; Warsaw University of Technology)

How to win MAI'2022 Depth Estimation challenge: manual backbone optimization combined with smart model pruning. (slides)
Łukasz Treszczotko (TCL Research Europe), Maciej Smyl (TCL Research Europe), Michał Kudelski (TCL Research Europe)

Interactive sequential analysis of a model improves the performance of human decision-making (slides)
Hubert Baniecki (University of Warsaw), Dariusz Parzych (Warsaw University of Technology), Przemysław Biecek (Warsaw University of Technology; University of Warsaw)

Interpretable Deep Learning with Prototypes
Bartosz Zieliński (Jagiellonian University, IDEAS NCBR)

Introduction and analysis of Diffusion-based Deep Generative Models (slides)
Kamil Deja (Warsaw University of Technology), Anna Kuzina (Vrije Universiteit Amsterdam), Tomasz Trzciński (Warsaw University of Technology, IDEAS NCBR, Tooploox, Jagiellonian University), Jakub Tomczak (Vrije Universiteit Amsterdam)

Lightweight Conditional Model Extrapolation for Streaming Data under Class Prior Shift (LIMES) (slides)
Paulina Tomaszewska (Warsaw University of Technology), Christoph H. Lampert (Institute of Science and Technology Austria)

NeRF – Generating 3d world from a bunch of images (slides)
Łukasz Pierścieniewski (NVIDIA), Jan Olszewski (NVIDIA)

One class classification approach to variational learning from Positive Unlabeled Data (slides)
Jan Mielniczuk (Polish Academy of Sciences; Warsaw University of Technology), Adam Wawrzeńczyk (Polish Academy of Sciences)

Improved Feature Importance Computation for Tree Models Based on the Banzhaf Value
Piotr Sankowski (IDEAS NCBR, MIM Solutions and University of Warsaw), Adam Karczmarz (IDEAS NCBR; University of Warsaw), Tomasz Michalak (IDEAS NCBR; University of Warsaw), Anish Mukherjee (IDEAS NCBR; University of Warsaw), Piotr Wygocki (MIM Solutions; University of Warsaw)

Reinforcement learning in Datacenters (slides)
Alessandro Seganti (Equinix), Darek Matlak (Equinix)

Sequential counterfactual prediction to support individualized decisions on treatment initiation
Paweł Morzywołek (Ghent University), Johan Steen (Ghent University), Wim Van Biesen (Ghent University), Johan Decruyenaere (Ghent University), Stijn Vansteelandt (Ghent University)

Harnessing the power of graph structures to understand changing user preferences over time and enhance item recommendations (slides)
Tomasz Palczewski (Samsung Research America), Anirudh Rao (Samsung Research America), Yingnan Zhu (Samsung Research America), Ayden Kim (Samsung Research America)

Towards self-certified learning: Probabilistic neural networks trained by PAC-Bayes with Backprop (slides)
Omar Rivasplata (University College London), María Pérez-Ortiz (University College London), John Shawe-Taylor (University College London), Csaba Szepesvári (University of Alberta, DeepMind Edmonton)

Uncertainty estimation in BERT-based Named Entity Recognition (slides)

Łukasz Rączkowski (Allegro Machine Learning Research), Riccardo Belluzzo (Allegro Machine Learning Research), Paweł Olszewski (Allegro Machine Learning Research), Piotr Zieliński (Allegro Machine Learning Research), Paweł Zawistowski (Allegro Machine Learning Research)

#Accepted Posters

Poster 01: survex: model-agnostic explainability for survival analysis (poster)
Mikołaj Spytek (MI2DataLab, Warsaw University of Technology), Mateusz Krzyziński (MI2DataLab, Warsaw University of Technology), Hubert Baniecki (MI2DataLab, University of Warsaw), Przemysław Biecek (MI2DataLab, Warsaw University of Technology)

Poster 02: forester: growing transparent tree-based models for everyone (poster)
Anna Kozak (Warsaw University of Technology), Hubert Ruczyński (Warsaw University of Technology), Patryk Słowakiewicz (Warsaw University of Technology), Adrianna Grudzień (Warsaw University of Technology)

Poster 03: Rapid reinforcement learning with the hierarchy of policies that chase fleeing targets
Michał Bortkiewicz (Warsaw University of Technology), Jakub Łyskawa, Artur Grudkowski, Paweł Wawrzyński, Mateusz Ostaszewski

Poster 04: Risk assessment of Polish joint stock companies using machine learning methods (poster)
Aleksandra Szymura (Wroclaw University of Economics and Business)

Poster 05: Informed Guided Rapidly-exploring Random Trees*-Connect for Near-Optimal Path Planning (poster)
Dominik Belter (Poznan University of Technology)

Poster 06: Continual learning on 3D point clouds with random compressed rehearsal (poster)
Maciej Zamorski (Wrocław University of Science and Technology, Hemolens Diagnostics), Michał Stypułkowski (University of Wrocław), Konrad Karanowski (Wrocław University of Science and Technology), Tomasz Trzciński (Warsaw University of Technology, IDEAS NCBR, Tooploox, Jagiellonian University), Maciej Zięba (Wrocław University of Science and Technology, Tooploox)

Poster 07: Machine Translation - from research to production (poster)
Karol Grzegorczyk (Allegro Machine Learning Research), Karol Karpiński (Allegro Machine Learning Reaserch), Mikołaj Koszowski (Allegro Machine Learning Reaserch)

Poster 08: Physics Informed Neural Network (PINN) for Infrasound propagation
Nikesh Yadav (Nicolaus Copernicus Astronomical Center, Polish Academy of Sciences)

Poster 09: Neural Architecture for Online Ensemble Continual Learning (poster)
Mateusz Wójcik (Alphamoon, Wroclaw University of Science and Technology), Witold Kościukiewicz (Alphamoon, Wroclaw University of Science and Technology), Adam Gonczarek (Alphamoon), Tomasz Kajdanowicz (Wroclaw University of Science and Technology)

Poster 10: QAOCDI: Deep Learning strikes at orthophotos analysis as never before (poster)
Przemyslaw Aszkowski (Poznan University of Technology), Bartosz Ptak (Poznan University of Technology), Marek Kraft (Poznan University of Technology)

Poster 11: STALKer: Terribly Accurate, Horribly Fast, Deep Learning Pipeline for Automatic Hemp Stalks Segmentation and Grasping Point Detection (poster)
Mateusz Piechocki (Poznan University of Technology), Dominik Pieczyński (Poznan University of Technology), Marek Kraft (Poznan University of Technology)

Poster 12: Contrastive Learning for Multilingual Named Entity Linking
Adam Kaczmarek (University of Wrocław)

Poster 13: Memory Enhanced Document-level Joint Entity and Relation Extraction (poster)
Witold Kościukiewicz (Alphamoon, Wroclaw University of Science and Technology), Mateusz Wójcik (Alphamoon, Wroclaw University of Science and Technology), Adam Gonczarek (Alphamoon), Tomasz Kajdanowicz (Wroclaw University of Science and Technology)

Poster 14: Across space and across seasons - aerial imagery place recognition for navigation support
Marek Kraft (Poznań University of Technology), Michał Węgrzynowski (Poznań University of Technology), Dominik Pieczyński (Poznań University of Technology)

Poster 15: Biased AI image generations models and misconceptions about health conditions, but not only
Adam Zadrożny (National Centre for Nuclear Research, University of Warsaw), Marianna Zadrożna (ASP)

Poster 16: How to see through the objects on the occluded scene? (poster)
Mikołaj Zieliński (Poznan University of Technology)

Poster 17: Parametrization of radiological reports using pattern matching and language models
Aleksander Obuchowski (Gdańsk University of Technology), Barbara Klaudel (Gdańsk University of Technology), Michał Stańczyk (SentiOne)

Poster 18: Automatic image generation for application of advertisements generation based on diffusion models (poster)
Karol Piniarski (Lingaro)

Poster 19: Automation of Citation Screening for Systematic Literature Reviews with the Usage of Eligibility Criteria
Wojciech Kusa (TU Wien), Allan Hanbury (TU Wien), Petr Knoth (The Open University)

Poster 20: Previous opinions is all you need. Legal opinions retrieval system (poster)
Maciej Osowski (Emplocity S.A.),Katarzyna Lorenc (Emplocity S.A.)

Poster 21: An active preference learning approach to aid the selection of validators in blockchain environments
Grzegorz Miebs (Poznan University of Technology; PSI Polska), Jonas Gehrlein (Web3 Foundation), Matteo Brunelli (University of Trento), Miłosz Kadziński (Poznan University of Technology)

Poster 22: Challenges of domain adaptation in medical domain (poster)
Barbara Klaudel (Radiato.ai; Gdańsk University of Technolog; Orange), Wasyl Badyra (Radiato.ai; Gdańsk University of Technology; Sii), Piotr Frąckowski (Radiato.ai; Gdańsk University of Technology; Enamor), Sebastian Krajna (Radiato.ai; Gdańsk University of Technology; MaxiEcu)

Poster 23: Coronary Ostia Localization Using Residual U-Net with Heatmap Matching and 3D DSNT (poster)
Miłosz Gajowczyk (Hemolens Diagnostics), Patryk Rygiel (Hemolens Diagnostics), Tomasz Konopczyński (Hemolens Diagnostics)

Poster 24: Review of the no-show problem (poster)
Magdalena Cebula (Orange), Jan Tarasiewicz, Mateusz Rymuza, Dominik Stępnień, Adam Skorek

Poster 25: Application of Machine Learning for Early Cancer Detection (poster)
Rohit Thanki (KRiAN GmbH), Purva Joshi (Wroclaw University of Science and Technology)

Poster 26: Deep learning based prediction of flow behind circular cylinders for application in wind Turbines
Nikesh Yadav (Nicolaus Copernicus Astronomical Center, Polish Academy of Sciences), Yurii Kurchenko (Warsaw University of Technology), Kacper Pawłowski (Warsaw University of Technology), Dominika Czerwińska (Warsaw University of Technology)

Poster 27: How could ML and AI help people with disabilities?
Anna Tumanova (Lingaro)

Poster 28: Monitoring system of NLP models - why? what? and how to built one yourself. (poster)
Wojciech Pyrak (Tidio), Łukasz Woronkiewicz (Tidio)

Poster 29: Neural-network based atomistic modeling of plastic deformation mechanisms in single-crystalline molybdenum (poster)
Amirhossein Naghdi Dorabati (NOMATEN CoE, National Centre for Nuclear Research), Mikko Alava (Aalto University), Stefanos Papanikolaou (NOMATEN CoE, National Centre for Nuclear Research)

Poster 30: Challenges in the automatic analysis of images from capsule endoscopy (poster)
Edyta Rogula (Biocam), Julia Pietraszek (Biocam), Amelia Okulewicz (Biocam)

Poster 31: Modern data augmentation, why so unintuitive? Overview of mixing augmentation methods. (poster)
Dominik Lewy (Lingaro)

Poster 32: State-of-the-Art Polish–Ukrainian Machine Translation (poster)
Helena Skowrońska (Shopai. sp. z o. o.), Paweł Wnuk (Shopai. sp. z o. o.)

Poster 33: BrightBox - explainable AI class tool to diagnose errors of ML models
Andżelika Zalewska (QED Software), Artur Żółkowski (QED Software, Warsaw University of Technology), Andrzej Janusz (QED Software)

Poster 34: Semantic Space of distributional models as compared to human’s abstract categorization and roots for creative association (poster)
Urszula Kuczma (Expertisly), Joanna Rączaszek-Leonardi (University of Warsaw), Kristian Tylen (University of Aarhus)