Research Projects

Do you prefer audio books over printed ones, and if so, what is the advantage? Why is video conferencing so awkward? Why do you rely on SMS messages as a means of communication in some situations but on e-mailing in others? Did you ever notice that you raise your eyebrows when asking people a question – even on the telephone? And did you ever wonder how to make your point in a discussion most convincingly?

We gradually increase our manual competence by exploring manual interaction spaces for many different kinds of objects. This is an active process that is very different from passive perception of "samples". The availability of humanoid robot hands offers the opportunity to investigate different strategies for such active exploration in realistic settings. In the present project, the investigation of such strategies shall be pursued from the perspective of „multimodal proprioception:“ correlating joint angles, partial contact information from touch sensors and joint torques as well as visual information about changes in finger and object position in such a way as to make predictions about "useful aspects" for shaping the ongoing interaction.

In the last decades, many researchers have used evolutionary algorithms to adapt the topology and connection weights of recurrent neural networks for various control tasks. This has become a useful machine learning technique. Because handling large genomes is difficult, however, these neural networks typically contain only a few neurons. If the genome contains a recipe for construction of the network instead of the network itself, it can be much smaller. We have developed a method than can exactly do this, and performs very well on a number of different problems.

Active Vision is a young and rapidly growing field in the area of image-analysis. It views vision as a process of active data aquisition.

Unlike most existing approaches to the grasp selection task for anthropomorphic robot hands, this vision-based project aims for a solution, which does not depend on an a-priori known 3D shape of the object. Instead it uses a decomposition of the object view (obtained from mono or stereo cameras) into local, grasping-relevant shape primitives, whose optimal grasp type and approach direction are known or learned beforehand. Based on this decomposition a list of possible grasps can be generated and ordered according to the anticipated overall grasp quality.

The spatio-temporal contact pattern during manipulation is a valuable source of information about object identity and object state, especially in uncertain environments. Using a bimanual robot manipulator setup with two 256 "pixel" touch sensor arrays, the present project is creating a "haptic pattern database" and investigates machine learning techniques to analyse the information contents of different haptic features and to extract identity and state information from haptic patterns. A closely connected goal are dynamic control strategies for contact movements with deformable or plastic objects, such as clay.

Together with the advances in building anthropomorphic robot hands, we are facing the question of how to dexterously control such complex robots with up to 20 degrees of freedom in up to five fingers and a wrist. Implementing fixed grasp and manipulation programs does not lead to satisfying results.
In our work, we propagate a manifold representation of such movements recorded from human demonstration. The main idea is to construct manifolds embedded in the finger joint angle space which represent the subspace of hand postures associated with a specific manipulation movement.

The goal of this project is to investigate the principles underlying co-evolution of a body shape and its neural controller. As a specific model system, we consider a robot hand that is controlled by a neural network. In contrast to existing work, we focus on the genetic regulation of neural circuits and morphological development. Our interest is directed at a better understanding of the facilitatory potential of co-evolution for the emergence of complex new functions, the interplay between development and evolution, the response of different genetic architectures to changing environments, as well as the role of important boundary constraints, such as wiring and tissue costs.

to be continued

to be continued

The ultimate goal of biological vision systems is to infer knowledge about the outside world that is relevant for the system in order to interact with its environment. Therefore, it is not sufficient to just determine the category or the mere object identity. Many variables of interest must be estimated, for example the distance towards an object, the size, orientation, velocity or even such abstract variables like the mood of another animal.

In this project we aim to extend existing invariant recognition approaches by using a new approach to hierarchical generative networks in order to implicitly represent (and learn) visual objects, and finally even scenes.