Research Projects

Machine Learning, Artificial Neural Networks and Kernel Methods
Evolutionary Algorithms
Autonomous Mobile Robots

Terminated projects (separate page).

Machine Learning, Artificial Neural Networks and Kernel Methods

Kernel-Based Multi-Target Drug Design Kernel-Based Multi-Target Drug Design - Development of multi-target models for virtual high-throughput screening and QSAR to validate the suitability of chemical compounds as potential drugs (A. Dörr, L. Rosenbaum).
Machine Learning Machine Learning for Document Security - Utilizing machine learning methods to detect malicious PDF and Flash documents and evaluating the security of those methods under adversarial conditions. (N. Šrndić)
ANTSAC ANTSAC is a new variant of the well-kwown RANdom SAmple Consensus (RANSAC) algorithm for robust estimation of model parameters. The idea of our method is based on the pheromone memory used in ant colony algorithms. (S. Otte)

Evolutionary Algorithms

JavaEvA EvA2 - a comprehensive Evolutionary Algorithms library: Hill-Climber,Simulated Annealing, Genetic Algorithms, Evolution Strategies,Genetic Programming, Grammatical Evolution, Model-Assisted ES, Multi-Objective ES (F. Becker).

Autonomous Mobile Robots

Terrain Mapping Terrain Mapping - We fuse data of a 3D LIDAR and a camera to classify terrain and obstacles in outdoor environments. With this information we build local terrain maps that can be used for localization and autonomous navigation. (S. Laible)
PATSY PATSY - (Person-recognizing Autonomous Transportation SYstem). Together with E&K Automation GmbH, Reutlingen, we develop the prototype of a transport system for medical containers in hospitals, which can recognize persons and other static and dynamic obstacles and plan/replan its route accordingly. (S. Buck, G. Rauscher, R. Hanten)
Outdoor Control cs::APEX - A framework for visual programming and data flow driven design that allows users to quickly prototype new algorithms by connecting inputs and outputs of computation nodes at run time. It currently is mainly used to prototype robot vision algorithms. (S. Buck, R. Hanten)
Outdoor Control Control of Outdoor Robots - Research on a quickly moving outdoor robot, which should be able to follow a fast moving person, while driving on an uneven terrain and avoiding both static and dynamic obstacles on the way. (G. Huskić)
3D Mouse 3D Mouse - A device which allows the user to manipulate virtual 3D Objects in all six degrees of freedom quickly and intuitively. (K. E. Wenzel, A. Masselli)
3D Mouse Surgical Navigation - In our research project we develop a navigation system for surgeons applying image processing based on a combination of preoperative CT data and laparoscopic images. (C. R. Pech)
SBC2015 DLR SpaceBot Cup 2015 - This year we are one of the 10 German universities that participate in the competition organized by the German Space Agency (DLR), namely the DLR SpaceBot Cup 2015. The theme of the competition is space exploration, for which the robots will have to explore a simulated planet surface. (G. Huskić, S. Buck, G. Rauscher, A. Zwiener, R. Hanten)
Person detection Person detection and person following - The ability to detect and recognize persons is an essential part of human-robot interaction. In our research we investigate machine-learning-based methods for detecting persons in 3D LIDAR data. (S. Buck, R. Hanten)
Efficient RGBD-SLAM for Autonomous Micro Aerial Vehicles - We investigate and develop efficient methods for Micro Aerial Vehicles (MAVs) equipped with RGBD-sensors to fly autonomously in previously unknown environments while building a map using only onboard processing power. (S. A. Scherer)
Multi-Camera SLAM for MAVs - We extend a monocular visual SLAM system to utilize multiple cameras without significant overlapping in their respective field of views, in order to achieve more robust pose tracking. We finally use the visual SLAM system to enable autonomous navigation of a MAV in unknown environments. (S. Yang)
Autonomous Flying Robots Tracking Vehicles with Autonomous Flying Robots - A method for self-localization and controlling the pose of MAVs. Our onboard solutions deal with limited payload capacity and little processing power but allow visual indoor and outdoor flights and carrier tracking.
Stereo Vision for Autonomous MAVs We investigate methods for using stereo vision to construct autonomous micro aerial vehicles (MAVs). This requires work on stereo vision as well as stereo-based self-localization and obstacle mapping. We have developed an MAV system, which can perform autonomous flight maneuvers using only stereo vision and inertial measurement data. (K. Schauwecker)
QuadrotorMPC - Compared to the classical control methods widely deployed on micro aerial vehicles, such as PID controllers, a series of advanced alternatives have recently been developed and applied to MAVs. In our lab, we implement one of the optimal control methods, model predictive control, on our quadrotor flying robots. (Y. Liu)
Automated RFID Inventory - Due to its high potential and recent presence in retail, warehousing and logistics, we investigate the capabilities of RFID-based inventory, localization and mapping with mobile robots. (A. Koch)
Path Following with RFID tags in Unknown Environments - We propose a new method employing a combination of RFID and odometry measurements for path following purposes. Our approach features a cost-effective method if the robot is equipped with an RFID reader for inventory tasks. (R. Liu)
RGB-D object recognition - We utilize a mobile service robot with a Microsoft Kinect RGB-D sensor as an identification system for objects by utilizing the additional depth information and derived features. (L. Jiang)
Dense Stereo Vision SLAM for MAVs - In this project we use a quadrocopter equipped with a stereo vision camera. We introduce a new dense stereo algorithm to estimate the depth maps and use the depth maps in a visual SLAM system. (R. Ait-Jellal)
Face Tracking and Pose Estimation - We present a real time algorithm for mobile robots to track human faces and estimate face poses accurately, even when humans move freely and far away from the camera or go through different illumination conditions in uncontrolled environments. (M. Vo)


VirtualLiver VirtualLiver- Subject of this BMBF funded project is to explore the physiology and functionality of the liver at different scales (cellular, intercellular, lobular and organ level) with the integration of these scales as an aim. Detoxification as a major task plays thereby an important role (S. Hoffmann, R. Keller).
VirtualLiver Nuclear receptor influenced human hepatocyte model - Development of a kinetic model of human hepatocytes (liver cells) influenced by different nuclear receptors. (S. Hoffmann)
Metabolic Networks Optimizing Metabolic Pathways - Reconstruction and Modeling of kinetic dependencies in metabolic pathways of E. coli and C. glutamicum (A. Dräger)
Diabetes Biomarkers Identification of biomarkers for type 2 diabetes - from mass spectrometry data using machine learning learning approaches. Mathematical modelling and classification, based on "differentially expressed" mass-profiles of individuals, who are at high risk to develop type 2 Diabetes mellitus. (L. Rosenbaum)
MARCAR IMI MARCAR: bioMARkers and molecular tumor classification for non-genotoxic CARcinogenesis.
The main objective of this EU-IMI funded project is merging new genetic and epigenetic data with the latest methods and technology to find new biomarkers for non-genotoxic carcinogenesis.
(J. Eichner, M. Römer)

Terminated Projects