no code implementations • 13 May 2024 • Daniel Bogdoll, Iramm Hamdard, Lukas Namgyu Rößler, Felix Geisler, Muhammed Bayram, Felix Wang, Jan Imhof, Miguel de Campos, Anushervon Tabarov, Yitian Yang, Hanno Gottschalk, J. Marius Zöllner
In this work, we propose AnoVox, the largest benchmark for ANOmaly detection in autonomous driving to date.
1 code implementation • 23 Apr 2024 • Niklas Wagner, Felix Mätzler, Samed R. Vossberg, Helen Schneider, Svetlana Pavlitska, J. Marius Zöllner
Using a small-scaled MaxViT-based model architecture, we evaluate the impact of discrete expression category labels in training with the continuous valence and arousal labels.
Ranked #1 on Dominance Estimation on EMOTIC
no code implementations • 6 Feb 2024 • Daniel Bogdoll, Jing Qin, Moritz Nekolla, Ahmed Abouelazm, Tim Joseph, J. Marius Zöllner
Reinforcement Learning is a highly active research field with promising advancements.
no code implementations • 1 Feb 2024 • Marc Uecker, J. Marius Zöllner
In this paper, we introduce a method for estimating blind spots for sensor setups of autonomous or automated vehicles and/or robotics applications.
no code implementations • 30 Nov 2023 • Daniel Grimm, Maximilian Zipfl, Felix Hertlein, Alexander Naumann, Jürgen Lüttin, Steffen Thoma, Stefan Schmid, Lavdim Halilaj, Achim Rettinger, J. Marius Zöllner
Precisely predicting the future trajectories of surrounding traffic participants is a crucial but challenging problem in autonomous driving, due to complex interactions between traffic agents, map context and traffic rules.
no code implementations • 27 Nov 2023 • Svetlana Pavlitska, Hannes Grolig, J. Marius Zöllner
Increasing the model capacity is a known approach to enhance the adversarial robustness of deep learning networks.
no code implementations • 20 Nov 2023 • Daniel Bogdoll, Yitian Yang, J. Marius Zöllner
Learning unsupervised world models for autonomous driving has the potential to improve the reasoning capabilities of today's systems dramatically.
1 code implementation • 30 Oct 2023 • Faris Janjoš, Marcel Hallgarten, Anthony Knittel, Maxim Dolgov, Andreas Zell, J. Marius Zöllner
We leverage recent advances in the space of the VAE, the foundation of the CVAE, which show that a simple change in the sampling procedure can greatly benefit performance.
no code implementations • 18 Oct 2023 • Abhishek Vivekanandan, Ahmed Abouelazm, Philip Schörner, J. Marius Zöllner
Accurately forecasting the motion of traffic actors is crucial for the deployment of autonomous vehicles at a large scale.
no code implementations • 18 Sep 2023 • Maximilian Zipfl, Moritz Jarosch, J. Marius Zöllner
Ensuring validation for highly automated driving poses significant obstacles to the widespread adoption of highly automated vehicles.
no code implementations • 18 Sep 2023 • Daniel Bogdoll, Svetlana Pavlitska, Simon Klaus, J. Marius Zöllner
Anomalies in the domain of autonomous driving are a major hindrance to the large-scale deployment of autonomous vehicles.
no code implementations • 5 Sep 2023 • Svetlana Pavlitska, Nico Lambing, Ashok Kumar Bangaru, J. Marius Zöllner
Real-time traffic light recognition is essential for autonomous driving.
no code implementations • 10 Aug 2023 • Daniel Bogdoll, Lukas Bosch, Tim Joseph, Helen Gremmelmaier, Yitian Yang, J. Marius Zöllner
We provide a characterization of world models and relate individual components to previous works in anomaly detection to facilitate further research in the field.
no code implementations • 17 Jul 2023 • Svetlana Pavlitska, Nico Lambing, J. Marius Zöllner
In this work, we survey existing works performing either digital or real-world attacks on traffic sign detection and classification models.
2 code implementations • 8 Jun 2023 • Faris Janjoš, Lars Rosenbaum, Maxim Dolgov, J. Marius Zöllner
The Variational Autoencoder (VAE) is a seminal approach in deep generative modeling with latent variables.
no code implementations • 6 Jun 2023 • Faris Janjoš, Max Keller, Maxim Dolgov, J. Marius Zöllner
Accurate vehicle trajectory prediction is an unsolved problem in autonomous driving with various open research questions.
no code implementations • 23 May 2023 • Ferdinand Mütsch, Helen Gremmelmaier, Nicolas Becker, Daniel Bogdoll, Marc René Zofka, J. Marius Zöllner
Simulation is an integral part in the process of developing autonomous vehicles and advantageous for training, validation, and verification of driving functions.
1 code implementation • 6 Feb 2023 • Daniel Bogdoll, Svenja Uhlemeyer, Kamil Kowol, J. Marius Zöllner
Deep neural networks (DNN) which are employed in perception systems for autonomous driving require a huge amount of data to train on, as they must reliably achieve high performance in all kinds of situations.
no code implementations • 24 Nov 2022 • Maximilian Zipfl, Moritz Jarosch, J. Marius Zöllner
Examining graphs for similarity is a well-known challenge, but one that is mandatory for grouping graphs together.
no code implementations • 26 Oct 2022 • Marc Uecker, Tobias Fleck, Marcel Pflugfelder, J. Marius Zöllner
To this end, we propose a novel computational taxonomy of LiDAR point cloud representations used in modern deep neural networks for 3D point cloud processing.
no code implementations • 27 Sep 2022 • Svetlana Pavlitskaya, Jonas Hendl, Sebastian Kleim, Leopold Müller, Fabian Wylczoch, J. Marius Zöllner
Adversarial patch-based attacks aim to fool a neural network with an intentionally generated noise, which is concentrated in a particular region of an input image.
no code implementations • 27 Sep 2022 • Svetlana Pavlitskaya, Joël Oswald, J. Marius Zöllner
Motivated by the fact that overfitted neural networks tend to rather memorize noise in the training data than generalize to unseen data, we examine how the training accuracy changes in the presence of increasing data perturbations and study the connection to overfitting.
no code implementations • 31 Aug 2022 • Larissa T. Triess, Christoph B. Rist, David Peter, J. Marius Zöllner
In a series of experiments, we demonstrate the application of our metric to determine the realism of generated LiDAR data and compare the realism estimation of our metric to the performance of a segmentation model.
no code implementations • 23 Aug 2022 • Svetlana Pavlitskaya, Nikolai Polley, Michael Weber, J. Marius Zöllner
In this work, we study whether temporal feature networks for object detection are vulnerable to universal adversarial attacks.
no code implementations • 15 Jul 2022 • Svetlana Pavlitskaya, Bianca-Marina Codău, J. Marius Zöllner
We have evaluated two approaches to generate naturalistic patches: by incorporating patch generation into the GAN training process and by using the pretrained GAN.
no code implementations • 13 Jul 2022 • Daniel Bogdoll, Meng Zhang, Maximilian Nitsche, J. Marius Zöllner
As a safety-critical problem, however, anomaly detection is a huge hurdle towards a large-scale deployment of autonomous vehicles in the real world.
no code implementations • 17 May 2022 • Barbara Schütt, Marc Heinrich, Sonja Marahrens, J. Marius Zöllner, Eric Sax
Verification and validation are major challenges for developing automated driving systems.
1 code implementation • 3 May 2022 • Daniel Bogdoll, Enrico Eisen, Maximilian Nitsche, Christin Scheib, J. Marius Zöllner
Tremendous progress in deep learning over the last years has led towards a future with autonomous vehicles on our roads.
no code implementations • 22 Apr 2022 • Svetlana Pavlitska, Christian Hubschneider, Lukas Struppek, J. Marius Zöllner
In this work, we apply sparse MoE layers to CNNs for computer vision tasks and analyze the resulting effect on model interpretability.
no code implementations • 21 Apr 2022 • Svetlana Pavlitskaya, Şiyar Yıkmış, J. Marius Zöllner
Coverage-guided testing (CGT) is an approach that applies mutation or fuzzing according to a predefined coverage metric to find inputs that cause misbehavior.
no code implementations • 9 Mar 2022 • Philipp Stegmaier, Karl Kurzer, J. Marius Zöllner
It can be demonstrated that the integration of risk metrics in the final selection policy consistently outperforms a baseline in uncertain environments, generating considerably safer trajectories.
no code implementations • 17 Feb 2022 • Larissa T. Triess, Andre Bühler, David Peter, Fabian B. Flohr, J. Marius Zöllner
Generative models can be used to synthesize 3D objects of high quality and diversity.
1 code implementation • 14 Feb 2022 • Karl Kurzer, Matthias Bitzer, J. Marius Zöllner
Cooperative trajectory planning methods for automated vehicles can solve traffic scenarios that require a high degree of cooperation between traffic participants.
1 code implementation • 7 Feb 2022 • Daniel Bogdoll, Moritz Nekolla, Tim Joseph, J. Marius Zöllner
Driving on roads is restricted by various traffic rules, aiming to ensure safety for all traffic participants.
1 code implementation • 3 Feb 2022 • Daniel Bogdoll, Felix Schreyer, J. Marius Zöllner
In this paper, we present ad-datasets, an online tool that provides such an overview for more than 150 data sets.
no code implementations • NeurIPS Workshop ICBINB 2021 • Larissa T. Triess, David Peter, J. Marius Zöllner
A number of applications, such as mobile robots or automated vehicles, use LiDAR sensors to obtain detailed information about their three-dimensional surroundings.
no code implementations • 19 Nov 2021 • Maximilian Zipfl, J. Marius Zöllner
Depending on the relative location between two traffic participants with respect to the road topology, semantically classified edges are created between the corresponding nodes.
1 code implementation • 5 Nov 2021 • Daniel Bogdoll, Johannes Jestram, Jonas Rauch, Christin Scheib, Moritz Wittig, J. Marius Zöllner
Due to the massive volume of sensor data that must be sent in real-time, highly efficient data compression is elementary to prevent an overload of network infrastructure.
no code implementations • 24 Sep 2021 • Larissa T. Triess, David Peter, Stefan A. Baur, J. Marius Zöllner
In a series of experiments, we confirm the soundness of our metric by applying it in controllable task setups and on unseen data.
no code implementations • 21 Sep 2021 • Faris Janjoš, Maxim Dolgov, J. Marius Zöllner
In this paper, we present a novel learned multi-modal trajectory prediction architecture for automated driving.
no code implementations • 20 Sep 2021 • Daniel Bogdoll, Jasmin Breitenstein, Florian Heidecker, Maarten Bieshaar, Bernhard Sick, Tim Fingscheidt, J. Marius Zöllner
Scaling the distribution of automated vehicles requires handling various unexpected and possibly dangerous situations, termed corner cases (CC).
no code implementations • 4 Jun 2021 • Larissa T. Triess, Mariella Dreissig, Christoph B. Rist, J. Marius Zöllner
Scalable systems for automated driving have to reliably cope with an open-world setting.
1 code implementation • 14 Apr 2021 • Moritz A. Zanger, Karam Daaboul, J. Marius Zöllner
Further, we provide theoretical and empirical analyses regarding the implications of model-usage on constrained policy optimization problems and introduce a practical algorithm that accelerates policy search with model-generated data.
no code implementations • 22 Mar 2021 • Michael Weber, Tassilo Wald, J. Marius Zöllner
For reliable environment perception, the use of temporal information is essential in some situations.
no code implementations • 12 Feb 2021 • Karl Kurzer, Philip Schörner, Alexander Albers, Hauke Thomsen, Karam Daaboul, J. Marius Zöllner
Data driven approaches for decision making applied to automated driving require appropriate generalization strategies, to ensure applicability to the world's variability.
no code implementations • 6 Apr 2020 • Larissa T. Triess, David Peter, Christoph B. Rist, J. Marius Zöllner
Autonomous vehicles need to have a semantic understanding of the three-dimensional world around them in order to reason about their environment.
1 code implementation • 30 Mar 2020 • Karl Kurzer, Christoph Hörtnagl, J. Marius Zöllner
Monte Carlo Tree Search (MCTS) has proven to be capable of solving challenging tasks in domains such as Go, chess and Atari.
no code implementations • 2 Feb 2020 • Karl Kurzer, Marcus Fechner, J. Marius Zöllner
Humans are well equipped with the capability to predict the actions of multiple interacting traffic participants and plan accordingly, without the need to directly communicate with others.
no code implementations • 28 Jun 2019 • Larissa T. Triess, David Peter, Christoph B. Rist, Markus Enzweiler, J. Marius Zöllner
This paper presents a novel CNN-based approach for synthesizing high-resolution LiDAR point cloud data.
no code implementations • 19 Apr 2019 • Michael Weber, Michael Fürst, J. Marius Zöllner
With automated focal loss we introduce a new loss function which substitutes this hyperparameter by a parameter that is automatically adapted during the training progress and controls the amount of focusing on hard training examples.
no code implementations • 3 Dec 2018 • Holger Banzhaf, Paul Sanzenbacher, Ulrich Baumann, J. Marius Zöllner
This paper introduces therefore a data-driven approach utilizing a deep convolutional neural network (CNN): Given the current driving situation, future ego-vehicle poses can be directly generated from the output of the CNN allowing to guide the motion planner efficiently towards the optimal solution.
1 code implementation • 10 Sep 2018 • Karl Kurzer, Florian Engelhorn, J. Marius Zöllner
Urban traffic scenarios often require a high degree of cooperation between traffic participants to ensure safety and efficiency.
no code implementations • 10 Sep 2018 • Peter Wolf, Karl Kurzer, Tobias Wingert, Florian Kuhnt, J. Marius Zöllner
This ensures a consistent model of the environment across scenarios as well as a behavior adaptation function, enabling on-line changes of desired behaviors without re-training.
no code implementations • 25 Jul 2018 • Karl Kurzer, Chenyang Zhou, J. Marius Zöllner
This work presents a Monte Carlo Tree Search (MCTS) based approach for decentralized cooperative planning using macro-actions for automated vehicles in heterogeneous environments.
no code implementations • 10 Sep 2017 • Florian Piewak, Timo Rehfeld, Michael Weber, J. Marius Zöllner
Grid maps are widely used in robotics to represent obstacles in the environment and differentiating dynamic objects from static infrastructure is essential for many practical applications.