Showing 1–2 of 2 results for author: Tülleners, I
-
A Semi-Automated Corner Case Detection and Evaluation Pipeline
Authors:
Isabelle Tulleners,
Tobias Moers,
Thomas Schulik,
Martin Sedlacek
Abstract:
In order to deploy automated vehicles to the public, it has to be proven that the vehicle can safely and robustly handle traffic in many different scenarios. One important component of automated vehicles is the perception system that captures and processes the environment around the vehicle. Perception systems require large datasets for training their deep neural network. Knowing which parts of th…
▽ More
In order to deploy automated vehicles to the public, it has to be proven that the vehicle can safely and robustly handle traffic in many different scenarios. One important component of automated vehicles is the perception system that captures and processes the environment around the vehicle. Perception systems require large datasets for training their deep neural network. Knowing which parts of the data in these datasets describe a corner case is an advantage during training or testing of the network. These corner cases describe situations that are rare and potentially challenging for the network. We propose a pipeline that converts collective expert knowledge descriptions into the extended KI Absicherung ontology. The ontology is used to describe scenes and scenarios that can be mapped to perception datasets. The corner cases can then be extracted from the datasets. In addition, the pipeline enables the evaluation of the detection networks against the extracted corner cases to measure their performance.
△ Less
Submitted 25 May, 2023;
originally announced May 2023.
-
Vision-Based Real-Time Indoor Positioning System for Multiple Vehicles
Authors:
Maximilian Kloock,
Patrick Scheffe,
Isabelle Tülleners,
Janis Maczijewski,
Stefan Kowalewski,
Bassam Alrifaee
Abstract:
We propose a novel external indoor positioning system that computes the position and orientation of multiple model-scale vehicles. For this purpose, we use a camera mounted at a height of 3.3m and LEDs attached to each vehicle. We reach an accuracy of about 1.1 cm for the position and around 0.6 ° for the orientation in the mean. Our system is real-time capable with a soft deadline of 20 ms. Moreo…
▽ More
We propose a novel external indoor positioning system that computes the position and orientation of multiple model-scale vehicles. For this purpose, we use a camera mounted at a height of 3.3m and LEDs attached to each vehicle. We reach an accuracy of about 1.1 cm for the position and around 0.6 ° for the orientation in the mean. Our system is real-time capable with a soft deadline of 20 ms. Moreover, it is robust against changing lighting conditions and reflections.
△ Less
Submitted 26 May, 2020; v1 submitted 13 February, 2020;
originally announced February 2020.
