no code implementations • 30 Jun 2021 • Stefan Zernetsch, Oliver Trupp, Viktor Kress, Konrad Doll, Bernhard Sick
This article presents a novel approach to incorporate visual cues from video-data from a wide-angle stereo camera system mounted at an urban intersection into the forecast of cyclist trajectories.
no code implementations • 4 Jun 2021 • Viktor Kress, Fabian Jeske, Stefan Zernetsch, Konrad Doll, Bernhard Sick
We compare our method with an approach that provides forecasts in the form of Gaussian distributions and discuss the respective advantages and disadvantages.
no code implementations • 19 Apr 2021 • Stefan Zernetsch, Hannes Reichert, Viktor Kress, Konrad Doll, Bernhard Sick
A basic movement detection based on motion history images (MHI) and a residual convolutional neural network (ResNet) are used to estimate probabilities for the current cyclist motion state.
no code implementations • 6 Mar 2018 • Stefan Zernetsch, Viktor Kress, Bernhard Sick, Konrad Doll
The method uses a deep Convolutional Neural Network (CNN) with a residual network architecture (ResNet), which is commonly used in image classification and detection tasks.