- Open Topics
AutoNOMOS CV-Lab Team
The CV-Lab's (Computer Vision) goal is to gather information out of a video stream. It's about the identification of lights, lanes and traffic signs, but also other vehicles, pedestrians and further objects. Having these objects verified by the insights of the 3D-Lab, precise assumptions on the state of the car's environment can be made.
Dr. Ernesto Tapia - TeamLead
Ernesto obtained a bachelor's degree from the Universidad Veracruzana in Mexico and a master in applied mathematics at the Centro de Investigación en Matemáticas (CIMAT). In 2004 he received a PhD degree in computer science from the Freie Universität Berlin. Since then he has been working there as a research assistant in the Artificial Intelligence Group.
He is currently working on self-localization and motion analysis from image data, automatic camera calibration, and generalization of image features for real-time object recognition. His scientific interests also include on-line recognition of handwritten mathematical notation, independence analysis and signal separation using neural networks, and image processing using mathematical morphology.
During his free time Ernesto likes cooking Mexican dishes and watching interesting movies.
Bennet Fischer - Research Assistant
Bennet joined the AI group in 2008 due to his work on humanoid robots. Since 2009 he is a member of the team AutoNOMOS.
He is working in the vision lab on the realization of computer vision algorithms inside hardware structures. Bennets main interests lie in the area between computer science and electrical engineering.
For recreation he enjoys the city on foot, via bike or his vintage Lancia.
Tobias Langner - Research Assistant
Tobias started to study computer science in 2003 at the Freie Universität Berlin with main focus on robotics and computer vision. Since 2009 he is a member of the FUmanoids, the humanoid robot soccer team of the FU Berlin, and graduated in 2010 with a thesis about robot self-localization.
His current work in the AutoNOMOS project comprehends of the detection of cars and pedestrians with classifiers based on Haar-like features. Furthermore he investigates more general approaches of obstacle detection using stereo vision and inverse perspective mapping.
Fritz Ulbrich - Research Assistant
Fritz began his studies in computer science at the Freie Universität Berlin in 2001. Since 2006 he is a member of the AutoNOMOS team. In 2009 he received his master's degree and since then he is a research assistant.
His main tasks are lane detection and the processing of the associated information. As test driver Fritz is frequently operating the autonomous cars 'Spirit of Berlin' and 'Made in Germany' and is often tweaking their sensor arrays.
After working hours Fritz likes to relax in a hammock with his laptop. Apart from that he enjoys to drive with his 'Schwalbe' moped through Berlin or to tinker with it.
Georg Bremer - Student Assistant
Georg studies computer science since 2004 at the Freie Universität Berlin with a short time stay at Universität Dortmund. He focuses mainly on pattern recognition and computer vision.
In the computer vision lab he works as student research assistent on brake light, taillight and turn signal detection. He also works on obstacle detection at the sides of the car and helped building a tool for creating a database to train a car detector.
Outside of his studies and work he likes to play with his daughter Fenja, if time allows cycling and tinkering on bicycles and everything else.
Robert Spangenberg - PhD Student
Robert studied computer science with a minor in economics from 2001 till mid 2007 at Humboldt-Universität zu Berlin. He wrote his diploma thesis at the Transport Research Institute of the German Aerospace Center. In 2007 he started to work at an automotive supplier, developing camera-based driver assistance systems. Since 2009 he is a part-time PhD student in the AutoNOMOS group.
Currently Robert's scientific interest lies in the usage of stereo cameras to improve the localisation of autonomous vehicles. This involves the usage of bundle adjustment and the evaluation and development of various interest point detectors to detect suitable landmarks. Earlier, he focused on Markov networks and particle filters in the context of multi-target tracking.
In his spare time he likes to do some free climbing, cycling or road running. Especially for climbing he has to travel a bit, but he likes that too.