Florian Bernard


Assistant Professor at University of Bonn

contact: f.bernardpi::gmail::com


10/2021: I joined the Department of Visual Computing at University of Bonn as Assistant Professor and head of the ‘Learning and Optimisation for Visual Computing’ group.

09/2021: Our paper Sparse Quadratic Optimisation over the Stiefel Manifold with Application to Permutation Synchronisation has been accepted at NeurIPS 2021.

05/2021: I received a CVPR 2021 Outstanding Reviewer Award.

03/2021: Two submissions were accepted and selected for oral presentation at CVPR 2021:
o Isometric Multi-Shape Matching
o i3DMM: Deep Implicit 3D Morphable Model of Human Heads

10/2020: Two ACM Transactions on Graphics (TOG) papers, to be presented at SIGGRAPH Asia, are now available:
o PIE: Portrait Image Embedding for Semantic Control
o RGB2Hands: Real-Time Tracking of 3D Hand Interactions from Monocular RGB Video

04/2020: I joined the Chair of Computer Vision & Artificial Intelligence at TU Munich as Visiting Professor.

News archive

Useful resources

News Archive

02/2020: Two of two submissions accepted at CVPR 2020 in Seattle:

01/2020: Our survey paper 3D Morphable Face Models - Past, Present and Future has been accepted (with minor revisions) for publication in the ACM Transactions on Graphics (TOG) journal.

07/2019: Our paper Higher-order Projected Power Iterations for Scalable Multi-Matching has been accepted for ICCV 2019 in Seoul.

03/2019: Our paper Synchronisation of Partial Multi-Matchings via Non-negative Factorisations has been accepted for publication in the Pattern Recognition journal.

02/2019: Two papers accepted for CVPR 2019 in Long Beach:

02/2019: Our paper Neural Animation and Reenactment of Human Actor Videos has been accepted (with minor revisions) for publication in the ACM Transactions on Graphics (TOG) journal.

26/09/2018: Our paper High-Fidelity Monocular Face Reconstruction based on an Unsupervised Model-based Face Autoencoder has been accepted for publication in the IEEE Transactions on Pattern Analysis and Machine Intelligence journal.

19/02/2018: Three out of three submission have been accepted for CVPR 2018 in Salt Lake City:

22/07/2017: Our paper MoFA: Model-based Deep Convolutional Face Autoencoder for Unsupervised Monocular Reconstruction has been accepted for ICCV 2017 in Venice.

14/03/2017: My PhD Thesis has been awarded the BVM-Award 2017.

27/02/2017: Our paper A Combinatorial Solution to Non-Rigid 3D Shape-to-Image Matching has been accepted for CVPR 2017 in Honolulu.


29/02/2016: Our paper Linear Shape Deformation Models with Local Support Using Graph-based Structured Matrix Factorisation has been accepted for CVPR 2016 in Las Vegas.

05/10/2015: Our paper Shape-aware 3D Interpolation using Statistical Shape Models has been awarded the Siemens/Sicas SHAPE Award (best paper).

05/03/2015: Our paper A Solution for Multi-Alignment by Transformation Synchronisation has been accepted for CVPR 2015 in Boston.