Analysis of the Effect of Various Input Representations for LSTM-Based Trajectory Prediction

Breuer, Antonia; Elflein, Sven; Joseph, Tim; Bolte, Jan-Aike; Homoceanu, Silviu; Fingscheidt, Tim

doi:10.1109/itsc.2019.8917373

Cited by 15 publications

(8 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Similar to our previous work [3], we define several distance metrics D T (j) , T (j) that can be used to evaluate the accuracy of a trajectory prediction algorithm. For a given object with index j, this distance metric D compares the predicted trajectory T (j) with the actual ground truth trajectory T (j) of the object, as given in the dataset.…”

Section: A Distance Metricsmentioning

confidence: 99%

“…Along the lines of our previous work [3], we propose to investigate the utility of the various information of the environment provided in the dataset. Thus, we invite parties interested in using the dataset to adopt their algorithms for trajectory prediction in such a way that they can work with variable input data.…”

Section: Challengesmentioning

confidence: 99%

“…, I} provided in the dataset is used by the trajectory prediction algorithm, we would like the authors to evaluate the benefit of the image. The image can be for example used as an input for a convolutional neural network (CNN), and be combined with a birds-eye view of the current traffic scene, implicitly encoding the structure of the surroundings [3].…”

Section: Challengesmentioning

confidence: 99%

“…Predicting the surrounding traffic scene is particularly difficult in urban and rural scenarios due to the high inter-dependency between the involved road users. One way to face this challenge is to use increasing amounts of data, causing a surge of popularity of data-driven approaches that rely on large-scale datasets in recent years [1], [2], [3], [4]. Other applications of trajectory datasets include the modeling and analysis of driving behavior [5] and the analysis of safety of the autonomous driving function [6].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

openDD: A Large-Scale Roundabout Drone Dataset

Breuer¹,

Termöhlen²,

Homoceanu³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Analyzing and predicting the traffic scene around the ego vehicle has been one of the key challenges in autonomous driving. Datasets including the trajectories of all road users present in a scene, as well as the underlying road topology are invaluable to analyze the behavior of the various traffic participants. The interaction between the traffic participants is especially high in intersection types that are not regulated by traffic lights, the most common one being the roundabout. We introduce the openDD dataset, including 84,774 accuratelytracked trajectories and HD map data of seven different roundabouts. The openDD dataset is annotated using images taken by a drone in 501 separate flights, totalling in over 62 hours of trajectory data. As of today the openDD is by far the largest publicly available trajectory dataset recorded from a drone perspective, while comparable datasets span 17 hours at most. The data is available, for both commercial and noncommercial use, at: http://www.l3pilot.eu/openDD.

show abstract

Section: A Distance Metricsmentioning

confidence: 99%

Section: Challengesmentioning

confidence: 99%

Section: Challengesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

openDD: A Large-Scale Roundabout Drone Dataset

Breuer¹,

Termöhlen²,

Homoceanu³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Therefore, RNNs have been used for the purpose of human motion prediction in different contexts [e.g. 9,10,11,12,13]. A common RNN approach is the usage of Long Short-Term Memory networks (LSTM).…”

Section: Introductionmentioning

confidence: 99%

Predicting Future Position From Natural Walking and Eye Movements with Machine Learning

Bremer

Stein

Lappe

2021

2021 IEEE International Conference on Artificial Intelligence and Virtual Reality (AIVR)

View full text Add to dashboard Cite

The prediction of human locomotion behavior is a complex task based on data from the given environment and the user. In this study, we trained multiple machine learning models to investigate if data from contemporary virtual reality hardware enables long-and short-term locomotion predictions. To create our data set, 18 participants walked through a virtual environment with different tasks. The recorded positional, orientation-and eye-tracking data was used to train an LSTM model predicting the future walking target. We distinguished between short-term predictions of 50ms and longterm predictions of 2.5 seconds. Moreover, we evaluated GRUs, sequence-to-sequence prediction, and Bayesian model weights. Our results showed that the best short-term model was the LSTM using positional and orientation data with a mean error of 5.14 mm. The best long-term model was the LSTM using positional, orientation and eye-tracking data with a mean error of 65.73 cm. Gaze data offered the greatest predictive utility for long-term predictions of short distances. Our findings indicate that an LSTM model can be used to predict walking paths in VR. Moreover, our results suggest that eye-tracking data provides an advantage for this task.

show abstract

Reducing Error Rate for Eye-Tracking System by Applying SVM

Ahmed

Nasrin

2022

Lecture Notes on Data Engineering and Communications Technologies

View full text Add to dashboard Cite

Analysis of the Effect of Various Input Representations for LSTM-Based Trajectory Prediction

Cited by 15 publications

References 12 publications

openDD: A Large-Scale Roundabout Drone Dataset

openDD: A Large-Scale Roundabout Drone Dataset

Predicting Future Position From Natural Walking and Eye Movements with Machine Learning

Reducing Error Rate for Eye-Tracking System by Applying SVM

Contact Info

Product

Resources

About