Goal-driven Self-Attentive Recurrent Networks for Trajectory Prediction

Chiara, Luigi Filippo; Coscia, Pasquale; Das, Sourav; Calderara, Simone; Cucchiara, Rita; Ballan, Lamberto

doi:10.1109/cvprw56347.2022.00282

Cited by 22 publications

(16 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Anchor-conditioned models use a set of anchor points to condition the prediction of an agent's trajectory [21,[26][27][28][29][30][31]. These anchor points can be, for example, the respective agent's estimated final goal position, which is used to condition the final trajectory estimation or even a trajectory proposal.…”

Section: Anchor-conditioned Methodsmentioning

confidence: 99%

“…While Dendorfer et al [30] modeled the multimodality of the task only through goal distribution estimation, which they argued eased the task of the trajectory decoder, Mangalam et al [28] sampled goals and waypoints from the predicted distributions and generated a conditioned probability distribution for every timestep of the trajectory. Chiara et al [26] followed this approach by sampling from goal distributions and injecting random noise into the trajectory generation module. While Mangalam et al [28] performed their computations completely in image space, Chiara et al [26] only predicted non-parametric probability distributions for possible goals in image space.…”

Section: Anchor-conditioned Methodsmentioning

confidence: 99%

“…Schmidt et al [111] showed that using navigational maps instead of HD maps by training a method with locally available HD maps and navigational maps via a teacher-student scheme is a viable alternative. In datasets like ETH/UCY, which do not offer maps but only images of the surroundings, methods integrate scene information into their trajectory predictions by encoding environment images using a convolutional neural network (CNN)-based backbone [16,26,28,39,79,96]. Incorporating environmental features (map information) improves the prediction performance as the models are able to learn traversable areas of the agents' surroundings [110].…”

Section: Possible Input Featuresmentioning

confidence: 99%

“…Many trajectory prediction methods utilize a simple scene representation strategy where the 2D position of agents is used without additional processing [20,26,69,74,102]. However, increasingly, researchers are exploring the use of graph-based scene representations.…”

Section: Scene Representation and Interaction Modelingmentioning

confidence: 99%

See 3 more Smart Citations

A Review of Trajectory Prediction Methods for the Vulnerable Road User

Schuetz,

Flohr

2023

Robotics

View full text Add to dashboard Cite

Predicting the trajectory of other road users, especially vulnerable road users (VRUs), is an important aspect of safety and planning efficiency for autonomous vehicles. With recent advances in Deep-Learning-based approaches in this field, physics- and classical Machine-Learning-based methods cannot exhibit competitive results compared to the former. Hence, this paper provides an extensive review of recent Deep-Learning-based methods in trajectory prediction for VRUs and autonomous driving in general. We review the state and context representations and architectural insights of selected methods, divided into categories according to their primary prediction scheme. Additionally, we summarize reported results on popular datasets for all methods presented in this review. The results show that conditional variational autoencoders achieve the best overall results on both pedestrian and autonomous driving datasets. Finally, we outline possible future research directions for the field of trajectory prediction in autonomous driving.

show abstract

Section: Anchor-conditioned Methodsmentioning

confidence: 99%

Section: Anchor-conditioned Methodsmentioning

confidence: 99%

Section: Possible Input Featuresmentioning

confidence: 99%

Section: Scene Representation and Interaction Modelingmentioning

confidence: 99%

See 2 more Smart Citations

A Review of Trajectory Prediction Methods for the Vulnerable Road User

Schuetz,

Flohr

2023

Robotics

View full text Add to dashboard Cite

show abstract

“…Recently, Transformer structure [63] is applied in this task [21,62,76,77] to model the spatio-temporal relations via an attention mechanism. Moreover, various viewpoints have emerged towards more practical applications, i.e., goal-driven idea [13,40,60,81], long-tail situation [39], interpretability [32], robustness [9,66,70,80], counterfactual analysis [11], planningdriven [12], generalization ability to new environment [6,27,72], and knowledge distillation [44].…”

Section: Trajectory Predictionmentioning

confidence: 99%

Uncovering the Missing Pattern: Unified Framework Towards Trajectory Imputation and Prediction

Xu¹,

Bazarjani²,

Chi³

et al. 2023

Preprint

View full text Add to dashboard Cite

Trajectory prediction is a crucial undertaking in understanding entity movement or human behavior from observed sequences. However, current methods often assume that the observed sequences are complete while ignoring the potential for missing values caused by object occlusion, scope limitation, sensor failure, etc. This limitation inevitably hinders the accuracy of trajectory prediction. To address this issue, our paper presents a unified framework, the Graph-based Conditional Variational Recurrent Neural Network (GC-VRNN), which can perform trajectory imputation and prediction simultaneously. Specifically, we introduce a novel Multi-Space Graph Neural Network (MS-GNN) that can extract spatial features from incomplete observations and leverage missing patterns. Additionally, we employ a Conditional VRNN with a specifically designed Temporal Decay (TD) module to capture temporal dependencies and temporal missing patterns in incomplete trajectories. The inclusion of the TD module allows for valuable information to be conveyed through the temporal flow. We also curate and benchmark three practical datasets for the joint problem of trajectory imputation and prediction. Extensive experiments verify the exceptional performance of our proposed method. As far as we know, this is the first work to address the lack of benchmarks and techniques for trajectory imputation and prediction in a unified manner.

show abstract

Trajectory Prediction with Heterogeneous Graph Neural Network

Luo

Yuan

et al. 2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Goal-driven Self-Attentive Recurrent Networks for Trajectory Prediction

Cited by 22 publications

References 31 publications

A Review of Trajectory Prediction Methods for the Vulnerable Road User

A Review of Trajectory Prediction Methods for the Vulnerable Road User

Uncovering the Missing Pattern: Unified Framework Towards Trajectory Imputation and Prediction

Trajectory Prediction with Heterogeneous Graph Neural Network

Contact Info

Product

Resources

About