Learning in the Curbside Coordinate Frame for a Transferable Pedestrian Trajectory Prediction Model

Jaipuria, Nikita; Habibi, Golnaz; How, Jonathan P.

doi:10.1109/itsc.2018.8569812

Cited by 13 publications

(13 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These methods do not assume any pedestrian dynamics, but instead, learn the dynamics from the observed data. A common approach is to cluster the trajectories from the observed data using Gaussian process [23] or vector fields [24] and learn the motion patterns. More recently, deep learning models [8], [25] have been developed to predict pedestrian trajectories using observed trajectories.…”

Section: B Pedestrian Trajectory Prediction Modelsmentioning

confidence: 99%

Analysis and Prediction of Pedestrian Crosswalk Behavior during Automated Vehicle Interactions

Jayaraman

Tilbury

Yang

et al. 2020

2020 IEEE International Conference on Robotics and Automation (ICRA)

View full text Add to dashboard Cite

For safe navigation around pedestrians, automated vehicles (AVs) need to plan their motion by accurately predicting pedestrians' trajectories over long time horizons. Current approaches to AV motion planning around crosswalks predict only for short time horizons (1-2 s) and are based on data from pedestrian interactions with human-driven vehicles (HDVs). In this paper, we develop a hybrid systems model that uses pedestrians' gap acceptance behavior and constant velocity dynamics for long-term pedestrian trajectory prediction when interacting with AVs. Results demonstrate the applicability of the model for long-term (> 5 s) pedestrian trajectory prediction at crosswalks. Further, we compared measures of pedestrian crossing behaviors in the immersive virtual environment (when interacting with AVs) to that in the real world (results of published studies of pedestrians interacting with HDVs), and found similarities between the two. These similarities demonstrate the applicability of the hybrid model of AV interactions developed from an immersive virtual environment (IVE) for real-world scenarios for both AVs and HDVs.

show abstract

Section: B Pedestrian Trajectory Prediction Modelsmentioning

confidence: 99%

Analysis and Prediction of Pedestrian Crosswalk Behavior during Automated Vehicle Interactions

Jayaraman

Tilbury

Yang

et al. 2020

2020 IEEE International Conference on Robotics and Automation (ICRA)

View full text Add to dashboard Cite

show abstract

“…In their method, trajectories are represented in terms of motion primitives and pair-wise transitions between them. Jaipuria et al [15] introduced Transferable Augmented Semi Non-negative Sparse Coding (TASNSC) as an extension of ASNSC [13]. In TASNSC, trajectories are mapped into a normalized, environment independent coordinate frame that helps learn a prediction model that can generalize to environments with different geometries.…”

Section: Related Workmentioning

confidence: 99%

“…This section introduces notations and building blocks for SILA, which is an incremental variation of ASNSC [13] and TASNSC [15].…”

Section: Background and Problem Statementmentioning

confidence: 99%

“…The motion primitives are learned using Augmented Semi Non-negative Sparse Coding (ASNSC) [13] and the pair-wise transitions between motion primitives are modeled using sparse GPs based on "pseudo inputs" [14]. We leverage the idea of a "curbside/common frame" as introduced in [15] to build a transferable model by mapping trajectory data into a normalized and environment independent common frame.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

SILA: An Incremental Learning Approach for Pedestrian Trajectory Prediction

Habibi

Jaipuria

How

2020

2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

Self Cite

View full text Add to dashboard Cite

The prediction of pedestrian motion is challenging, especially in crowded roads and intersections. Most of the current approaches apply offline methods to learn motion behaviors, but as a result, they are not able to learn continuously and typically do not generalize well to new environments. This paper presents Similarity-based Incremental Learning Algorithm (SILA) for pedestrian motion prediction with the ability of improving the learned model over the time as data is obtained incrementally. To keep the model size efficient, the motion primitives learned from the new data are compared with the previously known ones, and similar motion primitives are fused while novel motion primitives are added to the model. Results show that the SILA model growth rate is about 1/3 that of an incremental approach that does not fuse motion primitives. SILA is evaluated on different datasets and scenarios including intersections and busy streets. The results show that, even though SILA learns incrementally, it performs comparably to (and sometimes outperforms) state-of-the-art algorithms in pedestrian prediction. Additionally, SILA learning time only depends on the size of the data added incrementally, which makes SILA more efficient in terms of time and space compared to batch learning.

show abstract

“…Highly accurate vehicular trajectory estimation has become imminently important for automated vehicles and advanced driver assistance systems that have been developed in the recent years [1]. Accurate trajectory information has been used in various studies such as constraints for position estimation, 3-D mapping for automated vehicles, route planning, and vehicular control [2][3][4][5][6][7]. Accuracy is required for trajectory estimation of automated vehicles.…”

Section: Introductionmentioning

confidence: 99%

Vehicular trajectory estimation utilizing slip angle based on GNSS Doppler/IMU

et al. 2021

View full text Add to dashboard Cite

Accurate vehicular trajectory estimation is important for the recently developed autonomous driving systems. As the accuracy of the vehicular trajectory estimation is reduced with the slippage that occurs during turning, we propose a method in this study to accurately estimate the trajectory of a vehicle, focusing on the slip angle estimation. Although the two-wheel model is used as a general concept slip angle estimation, the accurate estimation of the parameters was difficult using the conventional methods. Therefore, a global navigation satellite system (GNSS) Doppler was used for parameter estimation. In addition, the roll angle was estimated as it occurs during turning and affects the slip angle of the vehicle. Specifically, we verified the improvement in accuracy of the vehicular trajectory estimation using the cost-effective GNSS Doppler/IMU.

show abstract

Learning in the Curbside Coordinate Frame for a Transferable Pedestrian Trajectory Prediction Model

Cited by 13 publications

References 16 publications

Analysis and Prediction of Pedestrian Crosswalk Behavior during Automated Vehicle Interactions

Analysis and Prediction of Pedestrian Crosswalk Behavior during Automated Vehicle Interactions

SILA: An Incremental Learning Approach for Pedestrian Trajectory Prediction

Vehicular trajectory estimation utilizing slip angle based on GNSS Doppler/IMU

Contact Info

Product

Resources

About