Federico Monti
Federico Monti
Federico Monti is a PhD student under the supervision of prof. Michael Bronstein, he moved to Università della Svizzera italiana in 2016 after achieving cum laude his B.Sc. and M.Sc. in Computer Science and Engineering at Politecnico di Milano. His research currently deals with the emerging field of Geometric Deep Learning and, in particular, with generalisation of Convolutional Neural Networks (CNNs) for signals defined on manifolds and graphs.GDL currently represents one of the latest and most active research fields in Machine Learning thanks to broad applicability and novelty it presents. Possible applications of GDL techniques range from Recommendation Systems (e.g. Matrix Completion problems), to Biomedical solutions (e.g. Disease Predictions), to Shape Analysis approaches (e.g. Shape Correspondence).
News
September, 2017
- Our paper "Geometric Matrix Completion with Recurrent Multi-Graph Neural Networks" has been accepted at NIPS2017.
July, 2017
- Intern at Google (July - October 2017).
June, 2017
- Invited speaker at the British Machine Vision Conference 2017 (“Deep Learning on Irregular Domains” workshop).
- Visiting Technical University of Munich (TUM).
May, 2017
- We present a new spectral graph CNN: CayleyNet. [PDF]
April, 2017
- New arXiv on "Geometric Matrix Completion with Recurrent Multi-Graph Neural Networks". [PDF]
February, 2017
- Reviewer for ICCV2017.
- Our paper "Geometric deep learning on graphs and manifolds using mixture model CNNs" (MoNet) has been accepted at CVPR2017 (oral presentation).
November, 2017
- A new spatial Convolutional Neural Network for graphs and manifolds: MoNet. [PDF]
Pubblications
CayleyNets: Graph Convolutional Neural Networks with Complex Rational Spectral Filters (2017)
Ron Levie*, Federico Monti*, Xavier Bresson, Michael M Bronstein
Geometric matrix completion with recurrent multi-graph neural networks (2017)
Federico Monti, Michael M. Bronstein, Xavier Bresson
Geometric deep learning on graphs and manifolds using mixture model CNNs (2016)
Federico Monti*, Davide Boscaini*, Jonathan Masci, Emanuele Rodolà, Jan Svoboda, Michael M. Bronstein
In Proc. CVPR2017 (oral presentation)
Deep convolutional neural networks for pedestrian detection (2015)
Denis Tomè*, Federico Monti*, Luca Baroffio, Luca Bondi, Marco Tagliasacchi, Stefano Tubaro
Journal of Signal Processing: Image Communication
