Video and 3D Analysis for Visual Learning

Starting Grant 279401 Videolearn

January 2012 - December 2016

Project members

Funded members:
Prof. Dr. Thomas Brox
Dr. Alexey Dosovitskiy
Dr. Margret Keuper
Nikolaus Mayer
Nima Sedaghat

Mohammadreza Zolfaghari
Benjamin Drayer
Gabriel Leivas Oliveira

Co-workers:
Maxim Tatarchenko
Benjamin Ummenhofer

Former members:
Philipp Fischer

A major challenge in computer vision these days is how to learn effective visual representations of the environment that can be used for decision making and planning. In this project, we focused particularly on learning representations from videos. We investigated how to obtain large amounts of annotated data and developed efficient interactive video segmentation solutions for manual annotation. Moreover, we found an effective way via automated rendering of images and annotation from synthetic scene models. We investigated how representations for motion and 3D structure can be learned end-to-end from training data and found very effective solutions that generalize to images that are very different from the training data. The project established deep learning methods for motion estimation, for 3D object modeling, and for two-frame structure from motion.

Publications enabled by this ERC Grant