About
38
Publications
15,964
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
754
Citations
Introduction
Stefan Ulbrich received the diploma degree in computer science from the University of Karlsruhe (TH), Germany in 2007. He received his PhD with his thesis on "Sensorimotor Learning for an Artificial Body Schema on Humanoid Robots" in 2014 and is currently a Postdoctoral Researcher at the Research Center for Information Technology at the Karlsruhe Institute of Technology (FZI). His major research interests are sensorimotor learning (i.e., body schema), spiking neural networks and motion capture.
Additional affiliations
October 2007 - December 2014
October 2007 - December 2014
March 2006 - September 2006
Education
October 2000 - July 2007
Publications
Publications (38)
The calibration of serial manipulators with high numbers of degrees of freedom by means of machine learning is a complex and time-consuming task. With the help of a simple strategy, this complexity can be drastically reduced and the speed of the learning procedure can be increased. When the robot is virtually divided into shorter kinematic chains,...
The previously presented Kinematic Bézier Maps (KBM) are a machine learning algorithm that has been tailored to efficiently learn the kinematics of redundant robots. This algorithm relies upon a representation based on projective geometry that uses a special set of polynomial functions borrowed from the field of Computer Aided Geometric Design (CAG...
ProSAT2 is a server to facilitate interactive visualization of sequence-based, residue-specific annotations mapped onto 3D
protein structures. As the successor of ProSAT (Protein Structure Annotation Tool), it includes its features for visualizing
SwissProt and PROSITE functional annotations. Currently, the ProSAT2 server can perform automated mapp...
Motion planning in the configuration space (C-space) induces benefits, such as smooth trajectories. It becomes more complex as the degrees of freedom (DOF) increase. This is due to the direct relation between the dimensionality of the search space and the DOF. Self-organizing neural networks (SONN) with their famous candidate, the Self-Organizing M...
We present a biologically inspired approach for path planning with dynamic obstacle avoidance. Path planning is performed in a condensed configuration space of a robot generated by self-organizing neural networks (SONN). The robot itself and static as well as dynamic obstacles are mapped from the Cartesian task space into the configuration space by...
Motion planning in the configuration space (C-space) induces benefits, such as smooth trajectories. It becomes more complex as the degrees of freedom (DOF) increase. This is due to the direct relation between the dimensionality of the search space and the DOF. Self-organizing neural networks (SONN) and their famous candidate, the Self-Organizing Ma...
For robotics, especially industrial applications, itis crucial to reactively plan safe motions through efficient algo-rithms. Planning is more powerful in the configuration spacethan the task space. However, for robots with many degreesof freedom, this is challenging and computationally expensive.Sophisticated techniques for motion planning such as...
Animal brains still outperform even the most performant machines with significantly lower speed. Nonetheless, impressive progress has been made in robotics in the areas of vision, motion- and path planning in the last decades. Brain-inspired Spiking Neural Networks (SNN) and the parallel hardware necessary to exploit their full potential have promi...
Without neuromorphic hardware, artificial stereo vision suffers from high resource demands and processing times impeding real-time capability. This is mainly caused by high frame rates, a quality feature for conventional cameras, generating large amounts of redundant data. Neuromorphic visual sensors generate less redundant and more relevant data s...
Safe, yet efficient, Human-robot interaction requires real-time-capable and flexible algorithms for robot control including the human as a dynamic obstacle. Even today, methods for collision-free motion planning are often computationally expensive, preventing real-time control. This leads to unnecessary standstills due to safety requirements. As na...
This chapter gives an in-depth overview of the activities and technical approaches of the team FLA²IR (FZI, MRK, OPEL) during the European Robotic Challenges to develop and deploy a flexible co-worker robot for the mounting of flexible polymer strips to a car door. The team won the EuRoC challenge with their extremely flexible approach that impleme...
Depth perception is crucial for many applications
including robotics, UAV and autonomous driving. The visual
sense, as well as cameras, map the 3D world on a 2D
representation, losing the dimension representing depth. A way
to recover 3D information from 2D images is to record and join
data from multiple viewpoints. In case of a stereo setup, 4D
da...
To guarantee safety in a shared work space between humans and robots, flexible robotic motion control is required. Unfortunately, path planning algorithms for complex robotic systems are too computationally expensive to enable a real-time solution on conventional hardware. With the long-term goal of performing a reactive path planning algorithm, we...
Understanding visual input as perceived by humans is a challenging task for machines. Today, most successful methods work by learning features from static images. Based on classical artificial neural networks, those methods are not adapted to process event streams as provided by the Dynamic Vision Sensor (DVS). Recently, an unsupervised learning ru...
Representation and execution of movement in biology is an active field of research relevant to neurorobotics. Humans can remember grasp motions and modify them during execution based on the shape and the intended interaction with objects. We present a hierarchical spiking neural network with a biologically inspired architecture for representing dif...
Artificial neural networks are known to perform function approximation but with increasingly large non-redundant input spaces, the number of required neurons grows drastically. Functions have to be sampled densely leading to large data sets which imposes problems for applications such as neurorobotics, and requires a long time for training. Further...
Combined efforts in the fields of neuroscience, computer science and biology allowed to design biologically realistic models of the brain based on spiking neural networks. For a proper validation of these models, an embodiment in a dynamic and rich sensory environment, where the model is exposed to a realistic sensory-motor task, is needed. Due to...
Bio-inspired robots still rely on classic robot control although advances in neurophysiology allow adaptation to control as well. However, the connection of a robot to spiking neuronal networks needs adjustments for each purpose and requires frequent adaptation during an iterative development. Existing approaches cannot bridge the gap between robot...
The ‘red-green’ pathway of the retina is classically recognized as one of the retinal mechanisms allowing humans to gather color information from light, by combining information from L-cones and M-cones in an opponent way. The precise retinal circuitry that allows the opponency process to occur is still uncertain, but it is known that signals from...
Objectives Even today's most advanced robots perform poorly at simple everyday tasks carried out routinely by humans and animals. This has very early motivated researchers to adopt neurobiological principles of cognition and control in robotics, yielding numerous approaches based on artificial neural networks and machine learning. However, many of...
Although robotics has made progress with respect to adaptability and interaction in natural environments, it cannot match the capabilities of biological systems. A promising approach to solve this problem is to create biologically plausible robot controllers that use detailed neuronal networks. However, this approach yields a large gap between the...
We present an extended version of our work on the design and implementation of a reference model of the human body, the Master Motor Map (MMM) which should serve as a unifying framework for capturing human motions, their representation in standard data structures and formats as well as their reproduction on humanoid robots. The MMM combines the def...
Software development plays a major role besides hardware setup and mechanical design when it comes to building complex robots such as mobile manipulators or humanoids. Different requirements have to be addressed depending on the application. A low-level controller for example must be implemented for realtime use, whereas a task planning component w...
In this work, we present a new software environment for the comparative evaluation of algorithms for grasping and dexterous manipulation. The key aspect in its development is to provide a tool that allows the reproduction of well-defined experiments in real-life scenarios in every laboratory and, hence, benchmarks that pave the way for objective co...
Simulation is essential for different robotic research fields such as mobile robotics, motion planning and grasp planning.
For grasping in particular, there are no software simulation packages, which provide a holistic environment that can deal
with the variety of aspects associated with this problem. These aspects include development and testing o...
This paper addresses the problem of hand-eye coordination and, more specifically, tool-eye recalibration of humanoid robots. Inspired by results from neuroscience, a novel method to learn the forward kinematics model as part of the body schema of humanoid robots is presented. By making extensive use of techniques borrowed from the field of computer...
Questions
Questions (2)
Imagine having a rough idea for an alternative building block of large language models other the well-known transformers.
In your head, the idea appears reasonable and overcome some perceived limitations of the state of the art. Training LLM, however, seems intimidating enough as Data Sets are well-hidden treasures, and fast-enough hardware is beyond reach for the majority of individual researchers.
How can one rapidly develop and evaluate a prototype while comparing to the established methods? Are there publicly available datasets that are small enough to allow for fast iterations (i.e., training without expensive hardware) yet meaningful enough to make at least an educated guess on how a model composed of these building blocks might behave at scale? Are there standardized benchmarks for this? Where would one begin starting this journey?
Hello
I was wondering if there is such a big enough data set available in oder to estimate the stochastic distribution of Covid 19-related symptoms in relation to location and time. This could be helpful for detecting variations/mutations in the disease and support diagnosis.
Thank you.