Continual learning is a machine learning domain that aims to mitigate catastrophic forgetting and enable models to be trained with an incoming stream of training data.
Multiple methods have been proposed in the past, yet one of the most promising groups of approaches leverage generative modelling to rehearse past data and retrain a model with a combination of current and regenerated samples.
dr hab. inż. Tomasz Trzciński will present an overview of generative models developed by the groups he is leading at Tooploox and Warsaw University of Technology. Tomasz is the Chief Scientist and co-owner at Tooploox and an Assistant Professor at Warsaw University of Technology and Jagiellonian University.
Obtained his Ph.D. in Computer Vision at École Polytechnique Fédérale de Lausanne in 2014. He has (co)-authored several papers in top-tier computer science conferences and high impact factor journals. He is an Associate Editor of IEEE Access and frequently serves as a reviewer in major computer vision conferences (CVPR, ICCV, ECCV, ACCV, BMVC, ICML, MICCAI) and international journals (TPAMI, IJCV, CVIU, TIP, TMM).
His professional appointments include work with Google in 2013, Qualcomm Corporate R&D in 2012 and Telefónica R&D in 2010. In 2017, he was appointed a Visiting Scholar at Stanford University. In 2016, he was named New Europe 100 Innovator as one of 100 outstanding challengers who are leading world-class innovation from Central and Eastern Europe.
He is co-organizing warsaw.ai – series of events focused on sharing expert AI knowledge, he is a member of IEEE and Computer Vision Foundation, as well as a member of Scientific Board for PLinML conference.
Digital marketer and copywriter experienced and specialized in AI, design, and digital marketing itself. Science, and holistic approach enthusiast, after-hours musician, and sometimes actor.