Room: Conference room 8

9:30- 9:40 Opening Remarks Taco Cohen, Qualcomm AI Research

9:40- 10:20 Graph neural networks and discrete latent structure in natural language processing Ivan Titov, University of Edinburgh

10:20- 11:00 Uncertainty quantification for manifold valued models Aasa Feragen, University of Copenhagen

Coffee Break

11:30- 12:10 Group Equivariant CNNs beyond Roto-Translations: B-Spline CNNs on Lie Groups Erik Bekkers, University of Amsterdam

12:10- 13:00 Discussion

Lunch Break

16:10- 16:50 TBA Risi Kondor, University of Chicago

16:50- 17:30 TBA Kostas Daniilidis, UPenn

Coffee Break

18:00- 18:40 Only Bayes Should Learn a Manifold Soren Hauberg, TU Denmark

18:40- 19:20 Regularizing Optimal Transport using Regularity Theory and Subspace Projections Marco Cuturi, Institut Polytechnique de Paris

19:30- 20:30 ELLIS Assembly