Physics-based scene understanding (01)

• Subject Area: Image and Language Processing, Computer Graphics and Visualisation, Human Computer Interaction, Ubiquitous and Wearable Computing
• Term: from 2017 to 2020
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 276693517

The goal of this project is to develop models that recover a physical interpretation of visual scenes. We aim to infer representations of images and videos that include semantic object properties and physical quantities such as weight, stiffness, shape, motion and others. This physical based understanding goes beyond what is currently possible with scene understanding models. We will develop new computational models based on generative models and deep learning that include meaningful and interpretable representations. These will be empirically validated on novel challenging datasets that will be created.

• DFG Programme: Collaborative Research Centres
• Subproject of: SFB 1233: Robust Vision - Inference Principles and Neural Mechanisms
• Applicant Institution: Eberhard Karls Universität Tübingen
• Project Heads: Dr. Peter Gehler, until 9/2017; Professor Dr.-Ing. Hendrik Lensch