Researchers send wake-up call to the car industry
A team of researchers in Tübingen show that optical flow systems based on deep neural networks – a likely component of future autonomous cars – are vulnerable to adversarial attacks. The computer vision experts are shaking up the automotive industry by warning car manufacturers around the globe that it could take a simple color pattern to put the brakes on computer vision systems in autonomous cars.
The Perceiving Systems Department, the Embodied Vision Group and the Autonomous Vision Group at the Max Planck Institute for Intelligent Systems are well represented at one of the world’s leading computer vision conferences taking place in Seoul in October.
The research group leader in the Perceiving Systems Department specializes in the development of computer-aided processes for the perception and digitisation of people and their activities in crowded scenes
From 30.6. to 7.7.2019 Application deadline: April 15, 2019
We are organizing the 3rd Computational Vision Summer School from 30.6. to 7.7.2019 in the beautiful black forest, Germany. It is aimed at PhD students and Postdocs with an interest in vision science research. Previous editions were fun, intellectually rewarding events for students and faculty alike. The summer school spans the entire spectrum of vision research from neuroscience and psychophysics to computer vision and fosters discussions at the intersection of biological and artificial vision systems. The school is unique in bringing together people from diverse disciplines in a very familiar atmosphere (only 60 people in total). CVSS is tuition free and accomodation is sponsored by the CRC/1233 Robust Vision.
Michael Black was awarded the University of British Columbia, Computer Science Department, Alumni Research Award 2018.
The German Association for Pattern Recognition praises her work to be “outstanding”
Tübingen – Dr. Siyu Tang received the DAGM MVTec 2018 Dissertation Award at the German Conference on Pattern Recognition (GCPR) conference in Stuttgart today. Her thesis with the title “People Detection and Tracking in Crowded Scenes” was praised by the board and the technical committee of the DAGM (Deutsche Arbeitsgemeinschaft für Mustererkennung) to be outstanding. “There was quite some competition this year”, said Reinhard Koch, the President of DAGM, which can be translated as German Association for Pattern Recognition. “This year we even decided to double the award money and grant two first places. Mrs. Tang will receive one of both. Congratulations for this achievement!” The award comes with a price money of 2.500 Euro and was given during the DAGM Award ceremony on Wednesday afternoon.
How do we perceive our own body weight? Which body weight do we find attractive?
An interdisciplinary team of scientists from the Max Planck Institute for Biological Cybernetics, the University of Tübingen and the Max Planck Institute for Intelligent Systems place test persons in front of their virtual selves and examine their self-perception. The aim of the studies is to investigate how accurately healthy women and men, incl. patients with anorexia nervosa, perceive their own body weight. The findings provide insights for new therapy approaches for people with eating disorders.
On the basis of a few photos alone, a new technique creates realistic avatars of animals that look and move like real animals
Filmmakers and developers of computer games will have a new way of animating animals in the future. A team led by researchers at the Max Planck Institute for Intelligent Systems in Tübingen, Germany, has developed a technique that uses photographs alone to create lifelike 3D models of almost all quadrupeds. These avatars can be animated to realistically imitate the movements of animals. But the simple method of bringing animals to life on the computer is not only interesting for the entertainment industry. Many people have lost a beloved family pet. Now this technology can bring them back to “life” as a virtual 3D avatar. It could also benefit biologists in species protection and help to make children in particular aware of the importance of biodiversity.
Am 1. März begann Geiger die Professur für "Learning-based Computer Vision" an der über 500 Jahre alten, renommierten Universität Tübingen. Geiger leitet weiterhin eine Forschungsgruppe am MPI-IS in Tübingen.