@incollection{LampertJungHubertetal.2022, author = {Lampert, Pascal and Jung, Janis and Hubert, Andreas and Doll, Konrad}, title = {Looping Through Color Space: A Simple Augmentation Method to Improve Biased Object Detection}, series = {Lecture Notes in Networks and Systems}, booktitle = {Lecture Notes in Networks and Systems}, publisher = {Springer Nature Singapore}, address = {Singapore}, isbn = {9789811916069}, issn = {2367-3370}, doi = {10.1007/978-981-19-1607-6_61}, pages = {687 -- 698}, year = {2022}, subject = {Objekterkennung}, language = {en} } @inproceedings{KressJungZernetschetal.2019, author = {Kreß, Viktor and Jung, Janis and Zernetsch, Stefan and Doll, Konrad and Sick, Bernhard}, title = {Pose Based Start Intention Detection of Cyclists}, publisher = {IEEE}, address = {Auckland}, doi = {10.1109/ITSC.2019.8917215}, pages = {2381 -- 2386}, year = {2019}, abstract = {In this work, we present a new approach for start intention detection of cyclists based on 3D human pose estimation to increase their safety in road traffic. Start intention detection is realized using sequences of frame-wise estimated 3D poses. The poses were obtained by image sequences recorded by a stereo camera mounted behind the windshield of a moving vehicle. For training and evaluation, a dataset with 206 starting cyclists was created in real traffic. We demonstrate the advantages of this approach by comparing it to an existing, solely head trajectory based method. In particular, we investigate the performance for different observed time horizons ranging from 0.12 s up to 1.0 s as inputs for the two methods. This is of special importance for the protection of cyclists in road traffic, as they often only become visible to approaching vehicles shortly before dangerous situations occur. With an input length of 1.0 s the solely head trajectory based approach detects starting motions on average 0.834 s after the first motion of the bicycle with an F1-score of 97.5 \%. The pose based approach outperforms these results by achieving the same F1-score 0.135 s earlier. The advantages of the pose based method become even more obvious with shorter input lengths. With an input length of 0.12 s, the head based approach achieves an F1-score of 93.5 \% after 2.37 s, while the same score is reached after 0.668 s using poses.}, subject = {Fahrerassistenzsystem}, language = {en} } @inproceedings{KressJungZernetschetal.2019, author = {Kreß, Viktor and Jung, Janis and Zernetsch, Stefan and Doll, Konrad and Sick, Bernhard}, title = {Start Intention Detection of Cyclists using an LSTM Network}, series = {INFORMATIK 2019: 50 Jahre Gesellschaft f{\"u}r Informatik - Informatik f{\"u}r Gesellschaft (Workshop-Beitr{\"a}ge)}, booktitle = {INFORMATIK 2019: 50 Jahre Gesellschaft f{\"u}r Informatik - Informatik f{\"u}r Gesellschaft (Workshop-Beitr{\"a}ge)}, editor = {Draude, Claude and Lange, Martin and Sick, Bernhard}, publisher = {Gesellschaft f{\"u}r Informatik e.V.}, address = {Bonn}, isbn = {978-3-88579-689-3}, issn = {1617-5468}, doi = {10.18420/inf2019_ws25}, pages = {219 -- 228}, year = {2019}, abstract = {In this article, we present an approach for start intention detection of cyclists based on their head trajectories. Therefore, we are using a network architecture based on Long Short-Term Memory (LSTM) cells, which is able to handle input sequences of different lengths. This is important because, for example, due to occlusions, cyclists often only become visible to approaching vehicles shortly before dangerous situations occur. Hence, the dependency of the results on the input sequence length is investigated. We use a dataset with 206 situations where cyclists were transitioning from waiting to moving that was recorded from a moving vehicle in inner-city traffic.With an input sequence length of 1.0 s we achieve an F1-score of 96.2\% on average 0.680 s after the first movement of the bicycle. We obtain similar results for sequence lengths down to 0.2 s. For shorter sequences, the results regarding the F1-score and the mean detection time deteriorate considerably.}, subject = {Fahrerassistenzsystem}, language = {en} } @inproceedings{KressJungZernetschetal.2018, author = {Kreß, Viktor and Jung, Janis and Zernetsch, Stefan and Doll, Konrad and Sick, Bernhard}, title = {Human Pose Estimation in Real Traffic Scenes}, series = {2018 IEEE Symposium Series on Computational Intelligence (SSCI)}, booktitle = {2018 IEEE Symposium Series on Computational Intelligence (SSCI)}, publisher = {IEEE}, address = {Bangalore, India}, doi = {10.1109/SSCI.2018.8628660}, pages = {518 -- 523}, year = {2018}, subject = {Autonomes Fahrzeug}, language = {en} } @inproceedings{JungHubertDolletal.2022, author = {Jung, Janis and Hubert, Andreas and Doll, Konrad and Kr{\"o}hn, Michael and Stadler, Jochen}, title = {Prozessinnovation}, series = {Wissenstransfer im Spannungsfeld von Autonomisierung und Fachkr{\"a}ftemangel, Tagungsband, 18. AALE-Konferenz, Pforzheim, 09.03.-11.03.2022}, booktitle = {Wissenstransfer im Spannungsfeld von Autonomisierung und Fachkr{\"a}ftemangel, Tagungsband, 18. AALE-Konferenz, Pforzheim, 09.03.-11.03.2022}, publisher = {Hochschule f{\"u}r Technik, Wirtschaft und Kultur Leipzig}, doi = {10.33968/2022.20}, year = {2022}, abstract = {Manuelle Montageprozesse sind nach wie vor unverzichtbar in vielen Bereichen der produzierenden Industrie. Vor allem die Qualit{\"a}tskontrolle, sowie das Einlernen neuer Mitarbeitenden stellen Betriebe durch die voranschreitende Digitalisierung vor neue Herausforderungen. Assistenzsysteme k{\"o}nnen hier helfen, die L{\"u}cke zwischen Anforderungen und Qualifikation zu {\"u}berbr{\"u}cken. Wir stellen einen Ansatz zur intelligenten Assistenz vor, welcher auf einer kamerabasierten Erkennung von Arbeitsabl{\"a}ufen mit Hilfe von Methoden des maschinellen Lernens beruht. Das Assistenzsystem erzeugt automatisiert Hilfsmaterial zur Unterst{\"u}tzung der Werkenden. Zus{\"a}tzlich zur Darstellung der technischen Aspekte, werden psychologische Aspekte, wie Akzeptanz und Motivation untersucht.}, subject = {Assistenzsystem}, language = {de} }