Page Not Found
Page not found. Your pixels are in another canvas.
Sitemap
A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.
Pages
About me
About me
Page not in menu
This is a page not in th emain menu
Posts
publications
From Prediction to Planning With Goal Conditioned Lane Graph Traversals
Published in preprint on arxiv.org, 2023
The field of motion prediction for automated driving has seen tremendous progress recently, bearing ever-more mighty neural network architectures. Leveraging these powerful models bears great potential for the closely related planning task. In this letter we propose a novel goal-conditioning method and show its potential to transform a state-of-the-art prediction model into a goal-directed planner. Our key insight is that conditioning prediction on a navigation goal at the behaviour level outperforms other widely adopted methods, with the additional benefit of increased model interpretability. We train our model on a large open-source dataset and show promising performance in a comprehensive benchmark.
Recommended citation: Hallgarten, M., Stoll, M., & Zell, A. (2023). From Prediction to Planning With Goal Conditioned Lane Graph Traversals. arXiv preprint arXiv:2302.07753. https://arxiv.org/abs/2302.07753
Stay on Track: A Frenet Wrapper to Overcome Off-road Trajectories in Vehicle Motion Prediction
Published in preprint on arxiv.org, 2023
Predicting the future motion of observed vehicles is a crucial en- abler for safe autonomous driving. The field of motion prediction has seen large progress recently with State-of-the-Art (SotA) models achieving impressive re- sults on large-scale public benchmarks. However, recent work revealed that learning-based methods are prone to predict off-road trajectories in challenging scenarios. These can be created by perturbing existing scenarios with additional turns in front of the target vehicle while the motion history is left unchanged. We argue that this indicates that SotA models do not consider the map information sufficiently and demonstrate how this can be solved, by representing model inputs and outputs in a Frenet frame defined by lane centreline sequences. To this end, we present a general wrapper that leverages a Frenet representation of the scene and that can be applied to SotA models without changing their architecture. We demonstrate the effectiveness of this approach in a comprehensive benchmark us- ing two SotA motion prediction models. Our experiments show that this reduces the off-road rate on challenging scenarios by more than 90%, without sacrificing average performance.
Recommended citation: Hallgarten, M., Kisa, I.,Stoll, M., & Zell, A. (2023). Stay on Track: A Frenet Wrapper to Overcome Off-road Trajectories in Vehicle Motion Prediction. arXiv preprint arXiv:arXiv:2306.00605. https://arxiv.org/abs/2306.00605
Parting with Misconceptions about Learning-based Vehicle Motion Planning
Published in preprint on arxiv.org, 2023
The release of nuPlan marks a new era in vehicle motion planning research, offering the first large-scale real-world dataset and evaluation schemes requiring both precise short-term planning and long-horizon ego-forecasting. Existing systems struggle to simultaneously meet both requirements. Indeed, we find that these tasks are fundamentally misaligned and should be addressed independently. We further assess the current state of closed-loop planning in the field, revealing the limitations of learning-based methods in complex real-world scenarios and the value of simple rule-based priors such as centerline selection through lane graph search algorithms. More surprisingly, for the open-loop sub-task, we observe that the best results are achieved when using only this centerline as scene context (\ie, ignoring all information regarding the map and other agents). Combining these insights, we propose an extremely simple and efficient planner which outperforms an extensive set of competitors, winning the nuPlan planning challenge 2023.
Recommended citation: Dauner, D., Hallgarten, M., Geiger, A. & Chitta, K. (2023). Parting with Misconceptions about Learning-based Vehicle Motion Planning. arXiv preprint arXiv:arXiv:2306.07962. https://arxiv.org/abs/2306.07962
Rethinking Integration of Prediction and Planning in Deep Learning-Based Automated Driving Systems: A Review
Published in preprint on arxiv.org, 2023
Automated driving has the potential to revolutionize personal, public, and freight mobility. Besides the enormous challenge of perception, i.e. accurately perceiving the environment using available sensor data, automated driving comprises planning a safe, comfortable, and efficient motion trajectory. To promote safety and progress, many works rely on modules that predict the future motion of surrounding traffic. Modular automated driving systems commonly handle prediction and planning as sequential separate tasks. While this accounts for the influence of surrounding traffic on the ego-vehicle, it fails to anticipate the reactions of traffic participants to the ego-vehicle’s behavior. Recent works suggest that integrating prediction and planning in an interdependent joint step is necessary to achieve safe, efficient, and comfortable driving. While various models implement such integrated systems, a comprehensive overview and theoretical understanding of different principles are lacking. We systematically review state-of-the-art deep learning-based prediction, planning, and integrated prediction and planning models. Different facets of the integration ranging from model architecture and model design to behavioral aspects are considered and related to each other. Moreover, we discuss the implications, strengths, and limitations of different integration methods. By pointing out research gaps, describing relevant future challenges, and highlighting trends in the research field, we identify promising directions for future research.
Recommended citation: Hagedorn, S., Hallgarten, M., Stoll, M. & Condurache, A. (2023). Rethinking Integration of Prediction and Planning in Deep Learning-Based Automated Driving Systems: A Review. arXiv preprint arXiv:2308.05731. https://arxiv.org/abs/2308.05731
Conditional Unscented Autoencoders for Trajectory Prediction
Published in preprint on arxiv.org, 2023
The CVAE is one of the most widely-used models in trajectory prediction for AD. It captures the interplay between a driving context and its ground-truth future into a probabilistic latent space and uses it to produce predictions. In this paper, we challenge key components of the CVAE. 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. We find that unscented sampling, which draws samples from any learned distribution in a deterministic manner, can naturally be better suited to trajectory prediction than potentially dangerous random sampling. We go further and offer additional improvements, including a more structured mixture latent space, as well as a novel, potentially more expressive way to do inference with CVAEs. We show wide applicability of our models by evaluating them on the INTERACTION prediction dataset, outperforming the state of the art, as well as at the task of image modeling on the CelebA dataset, outperforming the baseline vanilla CVAE.
Recommended citation: Janjoš, F., Hallgarten, M., Knittel, A., Dolgov, M., Zell, A., & Zöllner, J. M. (2023). Conditional Unscented Autoencoders for Trajectory Prediction. arXiv preprint arXiv:2310.19944. https://arxiv.org/abs/2310.19944
talks
Multimodal Prediction and Planning for Autonomous Driving
Published:
Autonomous driving is a rapidly evolving field that has the potential to transform transportation. Multimodal prediction and planning are crucial for enabling autonomous vehicles to navigate complex traffic scenarios safely and efficiently. In this talk an overview of recent advances in multimodal prediction and planning for autonomous driving is given.
From Prediction to Planning with Goal Conditioned Lane Graph Traversals
Published:
The field of motion prediction for automated driving has seen tremendous progress recently, bearing ever-more mighty neural network architectures. Leveraging these powerful models bears great potential for the closely related planning task. In this talk we propose a novel goal-conditioning method and show its potential to transform a state-of-the-art prediction model into a goal-directed planner. Our key insight is that conditioning prediction on a navigation goal at the behaviour level outperforms other widely adopted methods, with the additional benefit of increased model interpretability. We train our model on a large open-source dataset and show promising performance in a comprehensive benchmark.
1st Place in the 2023 nuPlan Planning Challenge
Published:
I am thrilled to announce that our team CS_Tu won the 2023 nuPlan Challenge! It was an amazing journey to team up with Daniel Dauner and Kashyap Chitta. I will be at the upcoming CVPR Workshop on End-to-End Autonomous Driving to present our work. I am excited to find out more about other teams‘ ideas and findings. If you’re attending CVPR, I would be happy to meet you in person and discuss our experiences!
teaching
Current Topics in Deep Neural Networks
Seminar, University of Tuebingen, Cognitive Sciences, 2021
This seminar aims to cover current topics in the field of deep learning such as Neural Style Transfer, Adversarial Robustness, Graph Neural Networks, Advanced Optimization Strategies, Neural Architecture Search etc.
Proseminar: Topics in Deep Neural Networks
Proeminar, University of Tuebingen, Cognitive Sciences, 2022
This seminar aims to cover basic concepts in the field of deep learning. A special focus is on recurrent neural networks, autoencoders, transfer learning, training strategies and generative adversarial networks.
Proseminar: Topics in Deep Neural Networks
Proeminar, University of Tuebingen, Cognitive Sciences, 2022
This seminar aims to cover basic concepts in the field of deep learning. A special focus is on different learning paradigms, optimization, recent advances in CNNs, deep learning frameworks and applications for deep neural networks.
Lecture: Introduction to Neural Networks
Lecture, University of Tuebingen, Cognitive Sciences, 2023
This lecture gives a short introduction to the biological foundations of neural networks, the most important algorithms of artificial neural networks and their underlying theory.