The inD Dataset: A Drone Dataset of Naturalistic Road User Trajectories at German Intersections

Bock, Julian; Krajewski, Robert J.; Moers, Tobias; Runde, Steffen; Vater, Lennart; Eckstein, Lutz

doi:10.48550/arxiv.1911.07602

Cited by 17 publications

(30 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Datasets must allow flexibility in choosing single or multiple prediction targets, in order to facilitate both singleagent and joint trajectory prediction. Therefore, we selected the inD [39] and the INTERACTION datasets [7] that provide joint tracked data over a whole recording. In both datasets, we generated individual samples for all agents except pedestrians and bicycles by extracting all 2.5+3s segments (2.5s past and 3s prediction recorded with 10Hz) with a 1.5s spacing between the samples.…”

Section: B Datasetsmentioning

confidence: 99%

StarNet: Joint Action-Space Prediction with Star Graphs and Implicit Global Frame Self-Attention

Janjoš¹,

Dolgov²,

Zöllner³

2021

Preprint

View full text Add to dashboard Cite

In this work, we present a novel multi-modal multi-agent trajectory prediction architecture, focusing on map and interaction modeling using graph representation. For the purposes of map modeling, we capture rich topological structure into vector-based star graphs, which enable an agent to directly attend to relevant regions along polylines that are used to represent the map. We denote this architecture StarNet, and integrate it in a single-agent prediction setting. As the main result, we extend this architecture to joint scene-level prediction, which produces multiple agents' predictions simultaneously.The key idea in joint-StarNet is integrating the awareness of one agent in its own reference frame with how it is perceived from the points of view of other agents. We achieve this via masked self-attention. Both proposed architectures are built on top of the action-space prediction framework introduced in our previous work, which ensures kinematically feasible trajectory predictions. We evaluate the methods on the interaction-rich inD and INTERACTION datasets, with both StarNet and joint-StarNet achieving improvements over state of the art.

show abstract

Section: B Datasetsmentioning

confidence: 99%

StarNet: Joint Action-Space Prediction with Star Graphs and Implicit Global Frame Self-Attention

Janjoš¹,

Dolgov²,

Zöllner³

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Action-to-action prediction requires acceleration and steering angle values in tracked data, otherwise they can be inferred from states via inverse models such as (5). Examples of datasets where accelerations are provided are [6], [36].…”

Section: B Learned Mappingsmentioning

confidence: 99%

“…We evaluated our approach on two real-world datasets recorded on German roads: inD (urban intersections) [6] and rounD (roundabouts) [36]. In total, they contain highly accurate drone-recorded trajectories of over 25000 agents of different classes, such as cars, trucks, buses, motorcycles, cyclists, and pedestrians.…”

Section: B Datasetsmentioning

confidence: 99%

“…IV-A) and SSP-ASP (Sec. IV-B) approaches on inD [6] and rounD [36] datasets. We benchmarked them against DKM [5], another method that achieves kinematically feasible predictions, albeit by using a state-to-action mapping introduced in Sec.…”

Section: Performancementioning

confidence: 99%

“…1. Multi-modal trajectory prediction (yellow) trained on the inD dataset [6], using the SSP-ASP approach Despite the large body of research, addressing vehicle trajectory prediction still remains a difficult problem to be solved especially for reasons such as complexity of understanding subtle social cues in multi-agent interactions and necessity of predicting multi-modal trajectories (see Fig. 1).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Self-Supervised Action-Space Prediction for Automated Driving

Janjoš¹,

Dolgov²,

Zöllner³

2021

Preprint

View full text Add to dashboard Cite

Making informed driving decisions requires reliable prediction of other vehicles' trajectories. In this paper, we present a novel learned multi-modal trajectory prediction architecture for automated driving. It achieves kinematically feasible predictions by casting the learning problem into the space of accelerations and steering angles -by performing action-space prediction, we can leverage valuable model knowledge. Additionally, the dimensionality of the action manifold is lower than that of the state manifold, whose intrinsically correlated states are more difficult to capture in a learned manner. For the purpose of action-space prediction, we present the simple Feed-Forward Action-Space Prediction (FFW-ASP) architecture. Then, we build on this notion and introduce the novel Self-Supervised Action-Space Prediction (SSP-ASP) architecture that outputs future environment context features in addition to trajectories. A key element in the self-supervised architecture is that, based on an observed action history and past context features, future context features are predicted prior to future trajectories. The proposed methods are evaluated on real-world datasets containing urban intersections and roundabouts, and show accurate predictions, outperforming state-of-the-art for kinematically feasible predictions in several prediction metrics.

show abstract

Haar Wavelet Based Block Autoregressive Flows for Trajectories

Bhattacharyya

Straehle

Fritz

et al. 2021

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Prediction of trajectories such as that of pedestrians is crucial to the performance of autonomous agents. While previous works have leveraged conditional generative models like GANs and VAEs for learning the likely future trajectories, accurately modeling the dependency structure of these multimodal distributions, particularly over long time horizons remains challenging. Normalizing flow based generative models can model complex distributions admitting exact inference. These include variants with split coupling invertible transformations that are easier to parallelize compared to their autoregressive counterparts. To this end, we introduce a novel Haar wavelet based block autoregressive model leveraging split couplings, conditioned on coarse trajectories obtained from Haar wavelet based transformations at different levels of granularity. This yields an exact inference method that models trajectories at different spatio-temporal resolutions in a hierarchical manner. We illustrate the advantages of our approach for generating diverse and accurate trajectories on two real-world datasets – Stanford Drone and Intersection Drone. Electronic supplementary material The online version of this chapter (10.1007/978-3-030-71278-5_20) contains supplementary material, which is available to authorized users.

show abstract

The inD Dataset: A Drone Dataset of Naturalistic Road User Trajectories at German Intersections

Cited by 17 publications

References 14 publications

StarNet: Joint Action-Space Prediction with Star Graphs and Implicit Global Frame Self-Attention

StarNet: Joint Action-Space Prediction with Star Graphs and Implicit Global Frame Self-Attention

Self-Supervised Action-Space Prediction for Automated Driving

Haar Wavelet Based Block Autoregressive Flows for Trajectories

Contact Info

Product

Resources

About