Project description
Towards a novel concept of dynamic scene understanding in computer vision
Computer vision lies at the intersection of computer science, mathematics, engineering and physics. Focussing on replicating parts of the complexity of the human vision system, it is one of the most powerful types of artificial intelligence. The EU-funded SIMULACRON project will develop a more profound concept of dynamic scene understanding in computer vision. It will focus on inferring the underlying physics (masses, elasticity, momenta and forces) and a simulation of the observed action directly from videos. Specifically, the project will develop algorithms for deformable shape modelling and variational methods for inferring physical simulations from videos. Ultimately, SIMULACRON will lead to a shift from 3D geometry inference to physical simulations inference.
Objective
Despite their amazing success, we believe that computer vision algorithms have only scratched the surface in terms of understanding our world from images. While most research on 3D reconstruction has been concerned with recovering the surface geometry and reflectance, SIMULACRON is focused on inferring the underlying physics (masses, elasticity, momenta, forces, etc.) and a simulation of the observed action directly from videos.
This not only provides a more profound understanding of the observed phenomena, but it also allows us to interpolate and extrapolate complex actions far beyond the observation: The inferred physical simulation can be employed for space-time super-resolution and for predictions into the future.
SIMULACRON covers three lines of research:
A) We will develop algorithms for deformable shape modeling. We will explore suitable representations of 3D shape and its evolution that enable the efficient computation of shape deformation, correspondence, interpolation and extrapolation. These techniques will form the basis for inferring physical simulations in parts B and C.
B) We will develop variational methods for inferring physical simulations from videos. We will compute a reference shape and simulation parameters that generate the shape deformation that is most consistent with the observations.
C) We will develop learning-based approaches for inferring physical simulations from videos. We will pursue two alternative approaches: First, we will generate synthetic training data by simulating deformable shapes and the associated camera observations. Second, we will devise self-supervised techniques for learning from real data without requiring labeled training data.
By shifting from inference of 3D geometry to inference of physical simulations, SIMULACRON will give rise to a more profound notion of dynamic scene understanding in computer vision, robotics and beyond. We believe that we have the necessary competence to pursue this project.
Fields of science
- engineering and technologyelectrical engineering, electronic engineering, information engineeringelectronic engineeringsensorsoptical sensors
- natural sciencesphysical sciencesopticsmicroscopysuper resolution microscopy
- natural sciencescomputer and information sciencesartificial intelligencecomputer vision
- natural sciencesmathematicspure mathematicsgeometry
- engineering and technologyelectrical engineering, electronic engineering, information engineeringelectronic engineeringrobotics
Keywords
Programme(s)
Topic(s)
Funding Scheme
ERC-ADG - Advanced GrantHost institution
80333 Muenchen
Germany