Master Thesis

Recovering High Frequency Motion Details for Dynamic Shapes

Work default illustration

Information

  • If you are interested in the proposal, please contact with the supervisors.

Description

In the last decade, many efforts have been made to jointly retrieve the 3D reconstruction of a deformable object captured by a single camera [1,2,3]. Nowadays, a wide number of solutions have been proposed for sparse [4,5], and even dense [6] observations. In this project, we will address the dense problem, focusing our formulation in order to recover complex non-rigid motion produced by high frequency actions that produces rapid time-varying details. To this end, we first review current techniques to finally present a novel method where different types of models and priors are exploited to code both global and local details in the observed scene. This master thesis will be carried on in the Institut de Robòtica i Informàtica Industrial, at the Universitat Politècnica de Catalunya (under direction of Dr. Antonio Agudo and Dr. Francesc Moreno).


Requisites:

Candidates with a background in mathematics, computer vision and good programming skills (Matlab/C++) are particularly encouraged to apply.


For additional information, please contact Dr. Antonio Agudo at aagudo@iri.upc.edu


[1] L. Torresani, A. Hertzmann and C. Bregler. Nonrigid structure from motion: estimating shape and motion with hierarchical priors. In TPAMI 30(5), 878-892, 2008.

[2] P.F.U. Gotardo and A. M. Martinez. Kernel non-rigid structure from motion with complementary rank-3 spaces. In ICCV, 2011.

[3] M. Lee, J. Cho, C. H. Choi and S. Oh. Procrustean normal distribution for non-rigid structure from motion. In CVPR, 2013.

[4] A. Agudo and F. Moreno-Noguer. Simultaneous pose and non-rigid shape with particle dynamics. In CVPR, 2015.

[5] A. Agudo and F. Moreno-Noguer. Learning shape, motion and elastic models in force space. In ICCV, 2015.

[6] R. Garg, A. Roussos and L. Agapito. Dense variational reconstruction of non-rigid surfaces from monocular video. In CVPR, 2013.

The work is under the scope of the following projects:

  • RobInstruct: Instructing robots using natural communication skills (web)