Multi-Agent Trajectory Prediction by Combining Egocentric and Allocentric Views

Xiaosong Jia, Liting Sun, Hang Zhao, Masayoshi Tomizuka, Wei Zhan
Proceedings of the 5th Conference on Robot Learning, PMLR 164:1434-1443, 2022.

Abstract

Trajectory prediction of road participants such as vehicles and pedestrians is crucial for autonomous driving. Recently, graph neural network (GNN) is widely adopted to capture the social interactions among the agents. Many GNN-based models formulate the prediction task as a single-agent prediction problem where multiple inference is needed for multi-agent prediction (which is common in practice), which leads to fundamental inconsistency in terms of homotopy as well as inefficiency for the memory and time. Moreover, even for models that do perform joint prediction, typically one centric agent is selected and all other agents’ information is normalized based on that. Such centric-only normalization leads to asymmetric encoding of different agents in GNN, which might harm its performance. In this work, we propose a efficient multi-agent prediction framework that can predict all agents’ trajectories jointly by normalizing and processing all agents’ information symmetrically and homogeneously with combined egocentirc and allocentric views. Experiments are conducted on two interaction-rich behavior datasets: INTERACTION (vehicles) and TrajNet++ (pedestrian). The results show that the proposed framework can significantly boost the inference speed of the GNN-based model for multi-agent prediction and achieve better performance. In the INTERACTION dataset’s challenge, the proposed model achieved the 1st place in the regular track and generalization track.

Cite this Paper

BibTeX
@InProceedings{pmlr-v164-jia22a, title = {Multi-Agent Trajectory Prediction by Combining Egocentric and Allocentric Views}, author = {Jia, Xiaosong and Sun, Liting and Zhao, Hang and Tomizuka, Masayoshi and Zhan, Wei}, booktitle = {Proceedings of the 5th Conference on Robot Learning}, pages = {1434--1443}, year = {2022}, editor = {Faust, Aleksandra and Hsu, David and Neumann, Gerhard}, volume = {164}, series = {Proceedings of Machine Learning Research}, month = {08--11 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v164/jia22a/jia22a.pdf}, url = {https://proceedings.mlr.press/v164/jia22a.html}, abstract = {Trajectory prediction of road participants such as vehicles and pedestrians is crucial for autonomous driving. Recently, graph neural network (GNN) is widely adopted to capture the social interactions among the agents. Many GNN-based models formulate the prediction task as a single-agent prediction problem where multiple inference is needed for multi-agent prediction (which is common in practice), which leads to fundamental inconsistency in terms of homotopy as well as inefficiency for the memory and time. Moreover, even for models that do perform joint prediction, typically one centric agent is selected and all other agents’ information is normalized based on that. Such centric-only normalization leads to asymmetric encoding of different agents in GNN, which might harm its performance. In this work, we propose a efficient multi-agent prediction framework that can predict all agents’ trajectories jointly by normalizing and processing all agents’ information symmetrically and homogeneously with combined egocentirc and allocentric views. Experiments are conducted on two interaction-rich behavior datasets: INTERACTION (vehicles) and TrajNet++ (pedestrian). The results show that the proposed framework can significantly boost the inference speed of the GNN-based model for multi-agent prediction and achieve better performance. In the INTERACTION dataset’s challenge, the proposed model achieved the 1st place in the regular track and generalization track.} }
Endnote
%0 Conference Paper %T Multi-Agent Trajectory Prediction by Combining Egocentric and Allocentric Views %A Xiaosong Jia %A Liting Sun %A Hang Zhao %A Masayoshi Tomizuka %A Wei Zhan %B Proceedings of the 5th Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2022 %E Aleksandra Faust %E David Hsu %E Gerhard Neumann %F pmlr-v164-jia22a %I PMLR %P 1434--1443 %U https://proceedings.mlr.press/v164/jia22a.html %V 164 %X Trajectory prediction of road participants such as vehicles and pedestrians is crucial for autonomous driving. Recently, graph neural network (GNN) is widely adopted to capture the social interactions among the agents. Many GNN-based models formulate the prediction task as a single-agent prediction problem where multiple inference is needed for multi-agent prediction (which is common in practice), which leads to fundamental inconsistency in terms of homotopy as well as inefficiency for the memory and time. Moreover, even for models that do perform joint prediction, typically one centric agent is selected and all other agents’ information is normalized based on that. Such centric-only normalization leads to asymmetric encoding of different agents in GNN, which might harm its performance. In this work, we propose a efficient multi-agent prediction framework that can predict all agents’ trajectories jointly by normalizing and processing all agents’ information symmetrically and homogeneously with combined egocentirc and allocentric views. Experiments are conducted on two interaction-rich behavior datasets: INTERACTION (vehicles) and TrajNet++ (pedestrian). The results show that the proposed framework can significantly boost the inference speed of the GNN-based model for multi-agent prediction and achieve better performance. In the INTERACTION dataset’s challenge, the proposed model achieved the 1st place in the regular track and generalization track.
APA
Jia, X., Sun, L., Zhao, H., Tomizuka, M. & Zhan, W.. (2022). Multi-Agent Trajectory Prediction by Combining Egocentric and Allocentric Views. Proceedings of the 5th Conference on Robot Learning, in Proceedings of Machine Learning Research 164:1434-1443 Available from https://proceedings.mlr.press/v164/jia22a.html.

Related Material