Heterogeneous-Agent Trajectory Forecasting Incorporating Class Uncertainty

Ivanovic, Boris; Lee, Kuan‐Hui; Tokmakov, Pavel; Wulfe, Blake; McAllister, Rowan; Gaidon, Adrien; Pavone, Marco

doi:10.48550/arxiv.2104.12446

Cited by 8 publications

(12 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The main concerns in graph-based architectures are the type of graph (e.g., undirected, digraph, heterogeneous, spatiotemporal, or others), and how to construct and extract features from them. The most common approach is to use graphs to represent only the spatial interaction between traffic participants [92], [129], [131], [132], [135]. For instance, in Diehl et al [131], Spectral Graph Convolutional Network (GCN) and Graph Attention Network (GAT) were evaluated.…”

Section: H Graphsmentioning

confidence: 99%

A Review on Intention-aware and Interaction-aware Trajectory Prediction for Autonomous Vehicles

Gomes¹,

Wolf²

2022

Preprint

View full text Add to dashboard Cite

This paper presents a literature review on Intention-aware and Interaction-aware Trajectory Prediction for Autonomous Vehicle, which covers primary studies since 2008. The research questions try to understand how intention/interaction improves prediction, and what are the techniques and technologies employed in trajectory prediction.<div>Moreover, this paper is under review in the IEEE Transactions on Intelligent Transportation Systems.</div>

show abstract

Section: H Graphsmentioning

confidence: 99%

A Review on Intention-aware and Interaction-aware Trajectory Prediction for Autonomous Vehicles

Gomes¹,

Wolf²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…This feature improves the performance of the network in many tasks that deals with sequence data and temporal dependences. However, Bi-LSTM has a higher computational cost than standard LSTM [40], [83], [91], [92].…”

Section: Bidirectional Ltsmmentioning

confidence: 99%

“…There are some variations of networks based on LSTM for trajectory prediction, such as Bidirectional LSTM [40], [92], Stacked LSTM [15], [101], Structural LSTM [15], and Spatio-Temporal LSTM [94]. Hou et al [15] proposed a hierarchical RNN using an encoder-decoder architecture with two-layered LSTMs (Stacked Long Short-Term Memory) to model the interaction of traffic participants on a highway scenario.…”

Section: Structural Ltsmmentioning

confidence: 99%

“…This approach allows the predictor to consider different characteristics in the movement of each class. There are other primary studies that applied graph neural network using spatio-temporal graphs [92], [130], [137].…”

Section: H Graphsmentioning

confidence: 99%

“…This modification allows greater control over the data generation process, since there will be a distribution for each possible value of the conditional variable [140]. In trajectory prediction, this technique can also estimate the distribution of future vehicle positions [92], [132], [141]. Figure 3 shows a general framework for trajectory prediction, where the conditional variable is the maneuver intention.…”

Section: Autoencodersmentioning

confidence: 99%

See 2 more Smart Citations

A Review on Intention-aware and Interaction-aware Trajectory Prediction for Autonomous Vehicles

Gomes¹,

Wolf²

2022

Preprint

View full text Add to dashboard Cite

show abstract

A Hierarchical Forecasting Model of Pedestrian Crossing Behavior for Autonomous Vehicle

Yang,

Pulgarin,

Herrmann

2024

IEEE Access

View full text Add to dashboard Cite

Simulation of pedestrians in shared spaces poses a significant challenge in autonomous driving virtual testing. The simulation pedestrian model can respond to autonomous vehicle behaviour changes. We present HFPM: a Hierarchical Forecasting Pedestrian Model to imitate pedestrian behaviour. The model has three layers: the dynamics model layer, the path planning layer, and the decision layer. In the dynamics model layer, an improved force model with the heading direction of the pedestrian is developed based on the Social Force Model, which can model pedestrian-pedestrian interaction. In the path planning layer, an Artificial Potential Field model is modified to plan a feasible path to the individual goals. The planning layer has a prediction module to predict the trajectory of vehicles on the road in order to choose the best route with no collision. The decision layer is a finite state machine with five states: the pedestrian can approach, walk, wait, run and reach the goal. The resulting HFPM model can produce more accurate simulation results than previously developed policy-based models, as demonstrated through qualitative and quantitative comparisons with a baseline pedestrian model obtained from the CITR data set.

show abstract

Heterogeneous-Agent Trajectory Forecasting Incorporating Class Uncertainty

Cited by 8 publications

References 0 publications

A Review on Intention-aware and Interaction-aware Trajectory Prediction for Autonomous Vehicles

A Review on Intention-aware and Interaction-aware Trajectory Prediction for Autonomous Vehicles

A Review on Intention-aware and Interaction-aware Trajectory Prediction for Autonomous Vehicles

A Hierarchical Forecasting Model of Pedestrian Crossing Behavior for Autonomous Vehicle

Contact Info

Product

Resources

About