I am a PhD student at the center for cognition and computation, department of computer science, Goethe University, Frankfurt. My research interest is focused on cognitive vision systems, with an emphasis on biologically inspired continuous learning and memory updates.
I assist in teaching the courses Machine Learning and Systems Engineering.
Anomaly detection in automotive scenarios
As I am working on continuous learning and memory in vision systems, a big interest is detecting anomalies and learning from them. Our recent work includes detecting state-transitions of rear-facing vehicle lamps (tail light, stop lamps). We submitted our work to IEEE 20th International Conference on Intelligent Transportation Systems. You can see the two datasets as well as supplementary material here: http://www.ccc.cs.uni-frankfurt.de/anomaly-detection-for-automotive-visual-signal-transition-estimation/.
I am working on creating large automotive datasets. The first iteration included a webcam, a GPS-receiver and an IMU unit mounted behind the rear-view mirror of the car connected to a laptop which has been able to grab images with 960×720 and 10 frames per second. Since Dec. 2016, I created an integrated Android application that is able to grab images with a resolution of 1920×1080 @ 10 fps (google play store: https://play.google.com/store/apps/details?id=de.weis.multisensor_grabber, github: https://github.com/TobiasWeis/android-multisensorgrabber-2) to ease recording sequences and scale up to multiple people. The backend consists of a SQL database and an annotation framework based on python and flask.
I am keen to understand and build anything that is related to Machine Learning and Electronics, in my private time I enjoy fast activities like Motorcycle Racing, Mountainbiking and Running. In 2016, I took part in the Baltic Sea Circle Rallye with my team /AFK – Away From Keyboard.
Department of computer science,
Robert-Mayer-Straße 10, D-60325 Frankfurt am Main.