@inproceedings{KoehlerGoldhammerBaueretal.2012, author = {K{\"o}hler, Sebastian and Goldhammer, Michael and Bauer, Sebastian and Doll, Konrad and Brunsmann, Ulrich and Dietmayer, Klaus}, title = {Early Detection of the Pedestrian's Intention to Cross the Street}, series = {15th International IEEE Conference on Intelligent Transportation Systems (ITSC 2012), Anchorage, Alaska, USA.}, booktitle = {15th International IEEE Conference on Intelligent Transportation Systems (ITSC 2012), Anchorage, Alaska, USA.}, doi = {10.1109/ITSC.2012.6338797}, pages = {1759 -- 1764}, year = {2012}, subject = {Fahrerassistenzsystem}, language = {en} } @inproceedings{GoldhammerKoehlerDolletal.2016, author = {Goldhammer, Michael and K{\"o}hler, Sebastian and Doll, Konrad and Sick, Bernhard}, title = {Track-Based Forecasting of Pedestrian Behavior by Polynomial Approximation and Multilayer Perceptrons}, series = {Intelligent Systems and Applications - Extended and Selected Results from the SAI Intelligent Systems Conference (IntelliSys) 2015}, booktitle = {Intelligent Systems and Applications - Extended and Selected Results from the SAI Intelligent Systems Conference (IntelliSys) 2015}, publisher = {Springer International Publishing}, isbn = {978-3-319-33386-1}, pages = {259 -- 279}, year = {2016}, abstract = {We present an approach for predicting continuous pedestrian trajectories over a time horizon of 2.5 s by means of polynomial least squares approximation and multilayer perceptron (MLP) artificial neural networks. The training data are gathered from 1075 real urban traffic scenes with uninstructed pedestrians including starting, stopping, walking and bending in. The polynomial approximation provides an extraction of the principal information of the underlying time series in the form of the polynomial coefficients. It is independent of sensor parameters such as cycle time and robust regarding noise. Approximation and prediction can be performed very efficiently. It only takes 35 ms on an Intel Core i7 CPU. Test results show 28\% lower prediction errors for starting scenes and 32\% for stopping scenes in comparison to applying a constant velocity movement model. Approaches based on MLP without polynomial input or Support Vector Regression (SVR) models as motion predictor are outperformed as well.}, subject = {Fußg{\"a}nger}, language = {en} } @inproceedings{GoldhammerBrunsmannDoll2012, author = {Goldhammer, Michael and Brunsmann, Ulrich and Doll, Konrad}, title = {Verkehrssicherheitsforschung: Bildverabeitung an intelligenten Kreuzungen}, series = {Messe-Exponat und Vortrag, Vision 2012, Stuttgart, 06.-08. November}, booktitle = {Messe-Exponat und Vortrag, Vision 2012, Stuttgart, 06.-08. November}, year = {2012}, subject = {Bildverarbeitung}, language = {de} } @inproceedings{GoldhammerStrigelMeissneretal.2012, author = {Goldhammer, Michael and Strigel, Elias and Meissner, Daniel and Brunsmann, Ulrich and Doll, Konrad and Dietmayer, Klaus}, title = {Cooperative Multi Sensor Network for Traffic Safety Applications at Intersections}, series = {15th International IEEE Conference on Intelligent Transportation Systems (ITSC 2012), Anchorage, Alaska, USA.}, booktitle = {15th International IEEE Conference on Intelligent Transportation Systems (ITSC 2012), Anchorage, Alaska, USA.}, doi = {10.1109/ITSC.2012.6338672}, pages = {1178 -- 1183}, year = {2012}, subject = {Fahrerassistenzsystem}, language = {en} } @inproceedings{GoldhammerHubertKoehleretal.2014, author = {Goldhammer, Michael and Hubert, Andreas and K{\"o}hler, Sebastian and Zindler, Klaus and Brunsmann, Ulrich and Doll, Konrad and Sick, Bernhard}, title = {Analysis on Termination of Pedestrians' Gait at Urban Intersections}, series = {Intelligent Transportation Systems (ITSC), 2014 IEEE 17th International Conference on}, booktitle = {Intelligent Transportation Systems (ITSC), 2014 IEEE 17th International Conference on}, publisher = {IEEE}, address = {Qingdao, China}, doi = {10.1109/ITSC.2014.6957947}, pages = {1758 -- 1763}, year = {2014}, subject = {Fahrerassistenzsystem}, language = {en} } @inproceedings{GoldhammerDollBrunsmannetal.2014, author = {Goldhammer, Michael and Doll, Konrad and Brunsmann, Ulrich and Gensler, Andre and Sick, Bernhard}, title = {Pedestrian's Trajectory Forecast in Public Traffic with Artificial Neural Networks}, series = {22nd International Conference on Pattern Recognition (ICPR 2014), Stockholm, Sweden}, booktitle = {22nd International Conference on Pattern Recognition (ICPR 2014), Stockholm, Sweden}, publisher = {IEEE}, isbn = {978-1-4799-5208-3}, issn = {1051-4651}, doi = {10.1109/ICPR.2014.704}, pages = {4110 -- 4115}, year = {2014}, subject = {Fahrerassistenzsystem}, language = {en} } @article{KoehlerGoldhammerBaueretal.2013, author = {K{\"o}hler, Sebastian and Goldhammer, Michael and Bauer, Sebastian and Zecha, Stephan and Doll, Konrad and Brunsmann, Ulrich and Dietmayer, Klaus}, title = {Stationary Detection of the Pedestrian's Intention at Intersections}, series = {IEEE Intelligent Transportation Systems Magazine}, volume = {2013}, journal = {IEEE Intelligent Transportation Systems Magazine}, number = {5}, issn = {1939-1390}, pages = {87 -- 99}, year = {2013}, subject = {Fahrerassistenzsystem}, language = {en} } @inproceedings{KoehlerGoldhammerZindleretal.2015, author = {K{\"o}hler, Sebastian and Goldhammer, Michael and Zindler, Klaus and Doll, Konrad and Dietmayer, Klaus}, title = {Stereo-Vision-Based Pedestrian's Intention Detection in a Moving Vehicle}, series = {Intelligent Transportation Systems (ITSC), 2015 IEEE 18th International Conference on}, booktitle = {Intelligent Transportation Systems (ITSC), 2015 IEEE 18th International Conference on}, doi = {10.1109/ITSC.2015.374}, pages = {2317 -- 2322}, year = {2015}, abstract = {We present a method to detect starting, stopping and bending in intentions of pedestrians from a moving vehicle based on stereo-vision. The method focuses on urban scenarios where these pedestrian movements are common and may result in critical situations. Pedestrian intentions are determined by means of an image-based motion contour histogram of oriented gradient descriptor. It is based on silhouettes gathered from stereo data and does not require any compensation of appearance changes resulting from the ego-motion of a vehicle. Nevertheless, it covers small movements indicating a pedestrian's intention. A linear support vector machine with probabilistic estimates is used for classification. We evaluated our method on the publicly available Daimler Pedestrian Path Prediction Benchmark Dataset containing detections of a stateof-the-art pedestrian detector. We detect a pedestrian's stopping intention from 125 ms to 500 ms before standing still within an accuracy range of 80\% to 100\%. Bending in is detected from 320 ms to 570 ms after a first visible lateral body movement in the same accuracy range. The intention to cross the road from standing still (starting) is detected 250 ms after the first visible motion and, therefore, within the first step with an accuracy of 100\%.}, subject = {Bildfolgenanalyse}, language = {en} } @inproceedings{GoldhammerKoehlerDolletal.2015, author = {Goldhammer, Michael and K{\"o}hler, Sebastian and Doll, Konrad and Sick, Bernhard}, title = {Camera Based Pedestrian Path Prediction by Means of Polynomial Least-squares Approximation and Multilayer Perceptron Neural Networks}, series = {SAI Intelligent Systems Conference (IntelliSys), 2015}, booktitle = {SAI Intelligent Systems Conference (IntelliSys), 2015}, publisher = {IEEE}, doi = {10.1109/IntelliSys.2015.7361171}, pages = {390 -- 399}, year = {2015}, abstract = {This paper provides a method to forecast pedestrian trajectories by means of polynomial least-squares approximation and multilayer perceptron artificial neural networks for traffic safety applications. The approach uses camera based head tracking as input data to predict a continuous trajectory for a 2.5 s future time horizon. Training and test is performed using 1075 recorded tracks of uninstructed pedestrians in common public traffic situations, including many challenging scenarios like starting, stopping and bending in. The neural network approach has the ability to handle these scenes by learning a single implicit movement model independent of a specific motion type. The polynomial approximation provides an extraction of the principal information of the underlying time series in the form of the polynomial coefficients, high independence of input data, e.g., sample rate, and additional noise resistance. Our test results show 24\% lower prediction errors for starting scenes and 29\% for stopping scenes in comparison to a constant velocity Kalman filter. Approaches using MLP without polynomial input and the usage of Support Vector}, subject = {Neuronales Netz}, language = {en} } @inproceedings{ZernetschKohnenGoldhammeretal.2016, author = {Zernetsch, Stefan and Kohnen, Sascha and Goldhammer, Michael and Doll, Konrad and Sick, Bernhard}, title = {Trajectory prediction of cyclists using a physical model and an artificial neural network}, series = {Intelligent Vehicles Symposium (IV), 2016}, booktitle = {Intelligent Vehicles Symposium (IV), 2016}, doi = {10.1109/IVS.2016.7535484}, pages = {833 -- 838}, year = {2016}, abstract = {This article presents two methods for predicting the trajectories of cyclists at an intersection and compares them to a Kalman Filter (KF) approach. The first method uses a physical model of cyclists to predict their future position. The second method is based on a polynomial least-squares approximation in combination with a multilayer perceptron artificial neural network and is able to predict the future position of cyclists independent of their motion type such as "Starting", "Stopping", "Waiting" or "Passing". To evaluate the performance of the methods, 566 tracks (394 for training, 172 for testing) of uninstructed cyclists were recorded at a public intersection using a wide angle stereo camera system and laser scanners. Using the tracks as input data, the future trajectory was predicted for a time horizon of 2.5 s. For starting motions the prediction using the physical model leads to 27\% more accurate positions than the KF approach for a forecast horizon of 2.5 s. The neural network shows a 34\% more accurate result for starting and stopping motions and a similar result for waiting and passing motions.}, subject = {Kalman-Filter}, language = {en} } @inproceedings{GoldhammerGerhardZernetschetal.2013, author = {Goldhammer, Michael and Gerhard, Matthias and Zernetsch, Stefan and Doll, Konrad and Brunsmann, Ulrich}, title = {Early Prediction of a Pedestrian's Trajectory at Intersections}, series = {16th International IEEE Conference on Intelligent Transportation Systems - (ITSC 2013), The Hague, The Netherlands}, booktitle = {16th International IEEE Conference on Intelligent Transportation Systems - (ITSC 2013), The Hague, The Netherlands}, publisher = {IEEE}, isbn = {978-1-4799-2914-6}, doi = {10.1109/ITSC.2013.6728239}, pages = {237 -- 242}, year = {2013}, subject = {Fahrerassistenzsystem}, language = {en} } @article{GoldhammerKoehlerZernetschetal.2019, author = {Goldhammer, Michael and K{\"o}hler, Sebastian and Zernetsch, Stefan and Doll, Konrad and Sick, Bernhard and Dietmayer, Klaus}, title = {Intentions of Vulnerable Road Users - Detection and Forecasting by Means of Machine Learning}, series = {IEEE Transactions on Intelligent Transportation Systems}, volume = {2020}, journal = {IEEE Transactions on Intelligent Transportation Systems}, number = {21 / 7}, doi = {10.1109/TITS.2019.2923319}, pages = {3035 -- 3045}, year = {2019}, subject = {Fahrerassistenzsystem}, language = {en} } @misc{DollKoehlerGoldhammeretal.2015, author = {Doll, Konrad and K{\"o}hler, Sebastian and Goldhammer, Michael and Brunsmann, Ulrich}, title = {Pedestrian Movement Modelling and Trajectory Prediction at Urban Intersections}, series = {International IEEE Conference on Intelligent Transportation Systems (ITSC 2015)}, journal = {International IEEE Conference on Intelligent Transportation Systems (ITSC 2015)}, address = {Las Palmas de Gran Canaria, Spain}, year = {2015}, abstract = {Vortrag}, subject = {Fahrerassistenzsystem}, language = {en} }