Semantics-STGCNN: A Semantics-Guided Spatial-Temporal Graph Convolutional Network for Multi-Class Trajectory Prediction

Ben Rainbow, Qianhui Men and Hubert P. H. Shum
Proceedings of the 2021 IEEE International Conference on Systems, Man, and Cybernetics (SMC), 2021

 Citation: 16#

Semantics-STGCNN: A Semantics-Guided Spatial-Temporal Graph Convolutional Network for Multi-Class Trajectory Prediction
# According to Google Scholar 2024

Abstract

Predicting the movement trajectories of multiple classes of road users in real-world scenarios is a challenging task due to the diverse trajectory patterns. While recent works of pedestrian trajectory prediction successfully modelled the influence of surrounding neighbours based on the relative distances, they are ineffective on multi-class trajectory prediction. This is because they ignore the impact of the implicit correlations between different types of road users on the trajectory to be predicted - for example, a nearby pedestrian has a different level of influence from a nearby car. In this paper, we propose to introduce class information into a graph convolutional neural network to better predict the trajectory of an individual. We embed the class labels of the surrounding objects into the label adjacency matrix (LAM), which is combined with the velocity-based adjacency matrix (VAM) comprised of the objects' velocity, thereby generating a semantics-guided graph adjacency (SAM). SAM effectively models semantic information with trainable parameters to automatically learn the embedded label features that will contribute to the fixed velocity-based trajectory. Such information of spatial and temporal dependencies is passed to a graph convolutional and temporal convolutional network to estimate the predicted trajectory distributions. We further propose new metrics, known as Average2 Displacement Error (aADE) and Average Final Displacement Error (aFDE), that assess network accuracy more accurately. We call our framework Semantics-STGCNN. It consistently shows superior performance to the state-of-the-arts in existing and the newly proposed metrics.


Downloads


YouTube


Cite This Research

Plain Text

Ben Rainbow, Qianhui Men and Hubert P. H. Shum, "Semantics-STGCNN: A Semantics-Guided Spatial-Temporal Graph Convolutional Network for Multi-Class Trajectory Prediction," in SMC '21: Proceedings of the 2021 IEEE International Conference on Systems, Man, and Cybernetics, pp. 2959-2966, Melbourne, Australia, IEEE, Oct 2021.

BibTeX

@inproceedings{rainbow21semantics,
 author={Rainbow, Ben and Men, Qianhui and Shum, Hubert P. H.},
 booktitle={Proceedings of the 2021 IEEE International Conference on Systems, Man, and Cybernetics},
 series={SMC '21},
 title={Semantics-STGCNN: A Semantics-Guided Spatial-Temporal Graph Convolutional Network for Multi-Class Trajectory Prediction},
 year={2021},
 month={10},
 pages={2959--2966},
 numpages={8},
 doi={10.1109/SMC52423.2021.9658781},
 issn={2959-2966},
 publisher={IEEE},
 location={Melbourne, Australia},
}

RIS

TY  - CONF
AU  - Rainbow, Ben
AU  - Men, Qianhui
AU  - Shum, Hubert P. H.
T2  - Proceedings of the 2021 IEEE International Conference on Systems, Man, and Cybernetics
TI  - Semantics-STGCNN: A Semantics-Guided Spatial-Temporal Graph Convolutional Network for Multi-Class Trajectory Prediction
PY  - 2021
Y1  - 10 2021
SP  - 2959
EP  - 2966
DO  - 10.1109/SMC52423.2021.9658781
SN  - 2959-2966
PB  - IEEE
ER  - 


Supporting Grants


Similar Research

Ruochen Li, Stamos Katsigiannis and Hubert P. H. Shum, "Multiclass-SGCN: Sparse Graph-Based Trajectory Prediction with Agent Class Embedding", Proceedings of the 2022 IEEE International Conference on Image Processing (ICIP), 2022
Qianhui Men and Hubert P. H. Shum, "PyTorch-Based Implementation of Label-Aware Graph Representation for Multi-Class Trajectory Prediction", Software Impacts (SIMPAC), 2021

 

Last updated on 6 October 2024
RSS Feed