Flows, mappings and shapes

Workshop theme

Computer vision and computer graphics have seen significant recent advances in the analysis of 3D shape through the tools of optimization and flows in shape space. Similarly, long standing problems such as image registration and optical flow are increasingly influencing and benefitting from the computational implementations and representations of diffeomorphic and other mappings. Nevertheless, there is a clear opportunity to further expand and intermix the influence of these methods throughout computer vision. On the computer vision side, this workshop will bring together practitioners working on the recovery of shape from the real word (shape from silhouette, single-image reconstruction), shape registration, and general image-to-image registration. The optimization methods employed range from large-scale non-convex techniques where high convergence rates are crucial in practice, to methods on shape spaces and manifolds, where second-order methods are a current hot topic. Shape and registration are already linked via the paradigm of flows, and the workshop is an opportunity to explore the unification of these frameworks with other developing methodologies in computer vision. Examples of such interactions include advanced optimization techniques applied to image enhancement and reconstruction; new image evolution paradigms like non-local means and Beltrami flows for image restoration; interactions with probability theory and statistics, powerful applications of Markov random fields, graphical models and stochastic grammars; the explosion of applications of compressive sensing and sparsity to a wide range of imaging problems; dynamical systems theory applied to video analysis and understanding; interactions with global analysis, geometric mechanics and complex geometry bringing new developments in shape analysis. Because of this flourishing of new methods arising from diverse mathematical backgrounds, it has become more and more challenging for a single researcher to embrace the full spectrum of computer vision fields. This makes the INI programme timely and critically important. This workshop will bring together the leading experts in image processing, computer vision and shape analysis to discuss the directions that have the potential to lead to the next big breakthroughs and their implications for other allied disciplines.

The application is closed

Timetable

Monday 11th December 2017
09:00 to 09:50	Registration	No Room Required
09:50 to 10:00	Welcome from David Abrahams (INI Director)	No Room Required
10:00 to 11:00	Nir Sochen Tel Aviv University Point correspondences in the functional map framework	Room 1
11:00 to 11:30	Morning Coffee	No Room Required
11:30 to 12:30	Olga Veksler University of Western Ontario Convexity and Other Shape priors for Single and Multiple Object Segmentation	Room 1
12:30 to 13:30	Lunch @ Wolfson Court	No Room Required
13:30 to 14:30	Emanuele Rodolà Università degli Studi di Roma La Sapienza Spectral approaches to partial deformable 3D shape correspondence	Room 1
14:30 to 15:30	Wei Zhu University of Alabama at Birmingham Euler's elastica based segmentation models and the fast algorithms	Room 1
15:30 to 16:00	Afternoon Tea	No Room Required
16:00 to 17:00	Andrew Zisserman University of Oxford 3D Shape Inference from Images using Deep Learning	Room 1
17:00 to 18:00	Welcome Wine Reception at INI	No Room Required

Tuesday 12th December 2017
09:00 to 10:00	Laurent Younes Johns Hopkins University Riemannian Diffeomorphic Mapping and Some Applications	Room 1
10:00 to 11:00	Lok Ming Lui Chinese University of Hong Kong Recent advances of Computational Quasiconformal Geometry in Imaging, Graphics and Visions	Room 1
11:00 to 11:30	Morning Coffee	No Room Required
11:30 to 12:30	Lourdes Agapito University College London Capturing 3D models of deformable objects from monocular sequences	Room 1
12:30 to 13:30	Lunch @ Wolfson Court	No Room Required
13:30 to 14:30	Maks Ovsjanikov École Polytechnique Efficient regularization of functional map computations	Room 1
14:30 to 15:30	Weihong Guo Simultaneous Image Segmentation and Registration and Applications	Room 1
15:30 to 16:00	Afternoon Tea	No Room Required
16:00 to 17:00	Carl Olsson Lund University Compact Rank Models and Optimization	Room 1

Wednesday 13th December 2017
09:00 to 10:00	Yuri Boykov University of Western Ontario Low-order graphical models for shapes and hierarchies in segmentation	Room 1
10:00 to 11:00	Tieyong Zeng Hong Kong Baptist University Two-Stage/Three-Stage Method for Image Segmentation	Room 1
11:00 to 11:30	Morning Coffee	No Room Required
11:30 to 12:30	Stephen Marsland Massey University Langevin equations for landmark image registration with uncertainty	Room 1
12:30 to 13:30	Lunch @ Wolfson Court	No Room Required
13:30 to 17:00	Free Afternoon	No Room Required
19:30 to 22:00	Formal Dinner at Christ's College	No Room Required

Thursday 14th December 2017
09:00 to 10:00	Peter Michor Universität Wien Soliton solutions for the elastic metric on spaces of curves	Room 1
10:00 to 11:00	Gabriel Peyre CNRS - Ecole Normale Superieure Paris Optimal Transport and Deep Generative Models	Room 1
11:00 to 11:30	Morning Coffee	No Room Required
11:30 to 12:30	Alex Bronstein Geometry and learning in 3D correspondence problems	Room 1
12:30 to 13:30	Lunch @ Wolfson Court	No Room Required
13:30 to 14:30	John Ball University of Oxford; Heriot-Watt University Nonlinear elasticity and image processing	Room 1
14:30 to 15:30	Christopher Zach Toshiba Research Europe Ltd When to lift (a function to higher dimensions) and when not	Room 1
15:30 to 16:00	Afternoon Tea	No Room Required
16:00 to 17:00	Darryl Holm Imperial College London Stochastic Metamorphosis in Imaging Science	Room 1

Friday 15th December 2017
09:00 to 10:00	Zachary Boyd University of California, Los Angeles Formulations of community detection in terms of total variation and surface tension	Room 1
10:00 to 11:00	Elaine Crooks Swansea University Compensated convexity, multiscale medial axis maps, and sharp regularity of the squared distance function	Room 1
11:00 to 11:30	Morning Coffee	No Room Required
11:30 to 12:30	François-Xavier Vialard An interpolating distance between Wasserstein and Fisher-Rao	Room 1
12:30 to 13:30	Lunch @ Wolfson Court	No Room Required

Variational methods and effective algorithms for imaging and vision

Organisers

Andrew Fitzgibbon Microsoft (UK)
Xue-Cheng Tai Hong Kong Baptist University
Laurent Younes Johns Hopkins University
Kewei Zhang University of Sussex

Speakers

Lourdes Agapito University College London
John Ball University of Oxford; Heriot-Watt University
Zachary Boyd University of California, Los Angeles
Yuri Boykov University of Western Ontario
Alex Bronstein
Elaine Crooks Swansea University
Weihong Guo
Darryl Holm Imperial College London
Lok Ming Lui Chinese University of Hong Kong
Stephen Marsland Massey University
Peter Michor Universität Wien
Carl Olsson Lund University
Maks Ovsjanikov École Polytechnique
Gabriel Peyre CNRS - Ecole Normale Superieure Paris
Emanuele Rodolà Università degli Studi di Roma La Sapienza
Nir Sochen Tel Aviv University
Olga Veksler University of Western Ontario
François-Xavier Vialard
Laurent Younes Johns Hopkins University
Christopher Zach Toshiba Research Europe Ltd
Tieyong Zeng Hong Kong Baptist University
Wei Zhu University of Alabama at Birmingham
Andrew Zisserman University of Oxford

Show All

Participants

Lourdes Agapito University College London
Daniel Alexander University College London
Simon Arridge University College London
John Aston University of Cambridge
Angelica I. Aviles-Rivero University of Cambridge
John Ball University of Oxford; Heriot-Watt University
Alexander Belyaev Heriot-Watt University
Martin Benning University of Cambridge
Marta Betcke University College London
Karthik Bharath University of Nottingham
Andreas Bock Imperial College London; Københavns Universitet (University of Copenhagen)
Zachary Boyd University of California, Los Angeles
Yuri Boykov University of Western Ontario
Alex Bronstein
Aurélie Bugeau Université de Bordeaux
Liam Burrows University of Liverpool
Ke Chen University of Liverpool
Veronica Corona University of Cambridge
Elaine Crooks Swansea University
Noemie Debroux Université de Rouen
Liangjian Deng Hong Kong Baptist University
Gunay Dogan National Institute of Standards and Technology
Matthias Ehrhardt University of Cambridge
Andrew Fitzgibbon Microsoft (UK)
Silvia Gazzola University of Bath
Joana Grah The Alan Turing Institute
Barbara Gris KTH - Royal Institute of Technology
Weihong Guo
Michael Hintermüller Humboldt-Universität zu Berlin
Jan Holland
Martin Holler University of Graz
Darryl Holm Imperial College London
Ian Jermyn Durham University
Meiyue Jiang Peking University
Nick Kingsbury University of Cambridge
Lisa Maria Kreusser University of Cambridge
Sebastian Kurtek Ohio State University
Lukas Lang University of Cambridge
Pan Liu Cambridge Centre for Analysis; Carnegie Mellon University
haixia Liu Hong Kong University of Science and Technology
Lok Ming Lui Chinese University of Hong Kong
Stephen Marsland Massey University
Peter Michor Universität Wien
James Nagy Emory University
Mila Nikolova CNRS (Centre national de la recherche scientifique); ENS de Cachan
Carl Olsson Lund University
Antonio Orlando Universidad Nacional de Tucuman
Maks Ovsjanikov École Polytechnique
Simone Parisotto Cambridge Centre for Analysis
Gabriel Peyre CNRS - Ecole Normale Superieure Paris
Joerg Polzehl Weierstraß-Institut für Angewandte Analysis und Stochastik
Erlend Skaldehaug Riis University of Cambridge
Emanuele Rodolà Università degli Studi di Roma La Sapienza
Carola Schonlieb University of Cambridge
Aamod Shanker University of California, Berkeley
Ferdia Sherry University of Cambridge
Nir Sochen Tel Aviv University
Jack Spencer University of Liverpool
Xue-Cheng Tai Hong Kong Baptist University
Anis Theljani Liverpool Centre for Mathematics in Healthcare
Matthew Thorpe University of Cambridge
Olga Veksler University of Western Ontario
François-Xavier Vialard
Benedikt Wirth Westfalische Wilhelms-Universitat Munster
Evren Yarman WesternGeco (Schlumberger)
Ke Yin Huazhong University of Science and Technology
Laurent Younes Johns Hopkins University
Christopher Zach Toshiba Research Europe Ltd
Tieyong Zeng Hong Kong Baptist University
Daoping Zhang University of Liverpool
Kewei Zhang University of Sussex
Wei Zhu University of Alabama at Birmingham
Andrew Zisserman University of Oxford
Ozan Öktem KTH - Royal Institute of Technology; Karolinska Institute