A Hierarchical Hybrid Learning Framework for Multi-agent Trajectory Prediction
Accurate and robust trajectory prediction of neighboring agents is critical for autonomous vehicles traversing in complex scenes. Most methods proposed in recent years are deep learning-based due to their strength in encoding complex interactions. However, unplausible predictions are often generated...
Gespeichert in:
Veröffentlicht in: | arXiv.org 2023-03 |
---|---|
Hauptverfasser: | , , , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
container_end_page | |
---|---|
container_issue | |
container_start_page | |
container_title | arXiv.org |
container_volume | |
creator | Jiao, Yujun Miao, Mingze Yin, Zhishuai Chunyuan Lei Zhu, Xu Nie, Linzhen Tao, Bo |
description | Accurate and robust trajectory prediction of neighboring agents is critical for autonomous vehicles traversing in complex scenes. Most methods proposed in recent years are deep learning-based due to their strength in encoding complex interactions. However, unplausible predictions are often generated since they rely heavily on past observations and cannot effectively capture the transient and contingency interactions from sparse samples. In this paper, we propose a hierarchical hybrid framework of deep learning (DL) and reinforcement learning (RL) for multi-agent trajectory prediction, to cope with the challenge of predicting motions shaped by multi-scale interactions. In the DL stage, the traffic scene is divided into multiple intermediate-scale heterogenous graphs based on which Transformer-style GNNs are adopted to encode heterogenous interactions at intermediate and global levels. In the RL stage, we divide the traffic scene into local sub-scenes utilizing the key future points predicted in the DL stage. To emulate the motion planning procedure so as to produce trajectory predictions, a Transformer-based Proximal Policy Optimization (PPO) incorporated with a vehicle kinematics model is devised to plan motions under the dominant influence of microscopic interactions. A multi-objective reward is designed to balance between agent-centric accuracy and scene-wise compatibility. Experimental results show that our proposal matches the state-of-the-arts on the Argoverse forecasting benchmark. It's also revealed by the visualized results that the hierarchical learning framework captures the multi-scale interactions and improves the feasibility and compliance of the predicted trajectories. |
format | Article |
fullrecord | <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_2790193181</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2790193181</sourcerecordid><originalsourceid>FETCH-proquest_journals_27901931813</originalsourceid><addsrcrecordid>eNqNyr0OgjAUQOHGxESivMNNnEn6IwKjMRIGNQ7spJaCRWz1UmJ4ex18AKczfGdGAi4Ei9IN5wsSDkNHKeXbhMexCMh5B4XRKFHdjJI9FNMVTQ1HLdEa20KO8qHfDu_QOITT2HsTyVZbDyXKTivvcIIL6toob5xdkXkj-0GHvy7JOj-U-yJ6onuNevBV50a0X6p4klGWCZYy8d_1AblJPis</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2790193181</pqid></control><display><type>article</type><title>A Hierarchical Hybrid Learning Framework for Multi-agent Trajectory Prediction</title><source>Free E- Journals</source><creator>Jiao, Yujun ; Miao, Mingze ; Yin, Zhishuai ; Chunyuan Lei ; Zhu, Xu ; Nie, Linzhen ; Tao, Bo</creator><creatorcontrib>Jiao, Yujun ; Miao, Mingze ; Yin, Zhishuai ; Chunyuan Lei ; Zhu, Xu ; Nie, Linzhen ; Tao, Bo</creatorcontrib><description>Accurate and robust trajectory prediction of neighboring agents is critical for autonomous vehicles traversing in complex scenes. Most methods proposed in recent years are deep learning-based due to their strength in encoding complex interactions. However, unplausible predictions are often generated since they rely heavily on past observations and cannot effectively capture the transient and contingency interactions from sparse samples. In this paper, we propose a hierarchical hybrid framework of deep learning (DL) and reinforcement learning (RL) for multi-agent trajectory prediction, to cope with the challenge of predicting motions shaped by multi-scale interactions. In the DL stage, the traffic scene is divided into multiple intermediate-scale heterogenous graphs based on which Transformer-style GNNs are adopted to encode heterogenous interactions at intermediate and global levels. In the RL stage, we divide the traffic scene into local sub-scenes utilizing the key future points predicted in the DL stage. To emulate the motion planning procedure so as to produce trajectory predictions, a Transformer-based Proximal Policy Optimization (PPO) incorporated with a vehicle kinematics model is devised to plan motions under the dominant influence of microscopic interactions. A multi-objective reward is designed to balance between agent-centric accuracy and scene-wise compatibility. Experimental results show that our proposal matches the state-of-the-arts on the Argoverse forecasting benchmark. It's also revealed by the visualized results that the hierarchical learning framework captures the multi-scale interactions and improves the feasibility and compliance of the predicted trajectories.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Contingency ; Deep learning ; Kinematics ; Motion planning ; Multiagent systems ; Optimization ; Predictions ; Trajectories ; Transformers</subject><ispartof>arXiv.org, 2023-03</ispartof><rights>2023. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>776,780</link.rule.ids></links><search><creatorcontrib>Jiao, Yujun</creatorcontrib><creatorcontrib>Miao, Mingze</creatorcontrib><creatorcontrib>Yin, Zhishuai</creatorcontrib><creatorcontrib>Chunyuan Lei</creatorcontrib><creatorcontrib>Zhu, Xu</creatorcontrib><creatorcontrib>Nie, Linzhen</creatorcontrib><creatorcontrib>Tao, Bo</creatorcontrib><title>A Hierarchical Hybrid Learning Framework for Multi-agent Trajectory Prediction</title><title>arXiv.org</title><description>Accurate and robust trajectory prediction of neighboring agents is critical for autonomous vehicles traversing in complex scenes. Most methods proposed in recent years are deep learning-based due to their strength in encoding complex interactions. However, unplausible predictions are often generated since they rely heavily on past observations and cannot effectively capture the transient and contingency interactions from sparse samples. In this paper, we propose a hierarchical hybrid framework of deep learning (DL) and reinforcement learning (RL) for multi-agent trajectory prediction, to cope with the challenge of predicting motions shaped by multi-scale interactions. In the DL stage, the traffic scene is divided into multiple intermediate-scale heterogenous graphs based on which Transformer-style GNNs are adopted to encode heterogenous interactions at intermediate and global levels. In the RL stage, we divide the traffic scene into local sub-scenes utilizing the key future points predicted in the DL stage. To emulate the motion planning procedure so as to produce trajectory predictions, a Transformer-based Proximal Policy Optimization (PPO) incorporated with a vehicle kinematics model is devised to plan motions under the dominant influence of microscopic interactions. A multi-objective reward is designed to balance between agent-centric accuracy and scene-wise compatibility. Experimental results show that our proposal matches the state-of-the-arts on the Argoverse forecasting benchmark. It's also revealed by the visualized results that the hierarchical learning framework captures the multi-scale interactions and improves the feasibility and compliance of the predicted trajectories.</description><subject>Contingency</subject><subject>Deep learning</subject><subject>Kinematics</subject><subject>Motion planning</subject><subject>Multiagent systems</subject><subject>Optimization</subject><subject>Predictions</subject><subject>Trajectories</subject><subject>Transformers</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>BENPR</sourceid><recordid>eNqNyr0OgjAUQOHGxESivMNNnEn6IwKjMRIGNQ7spJaCRWz1UmJ4ex18AKczfGdGAi4Ei9IN5wsSDkNHKeXbhMexCMh5B4XRKFHdjJI9FNMVTQ1HLdEa20KO8qHfDu_QOITT2HsTyVZbDyXKTivvcIIL6toob5xdkXkj-0GHvy7JOj-U-yJ6onuNevBV50a0X6p4klGWCZYy8d_1AblJPis</recordid><startdate>20230324</startdate><enddate>20230324</enddate><creator>Jiao, Yujun</creator><creator>Miao, Mingze</creator><creator>Yin, Zhishuai</creator><creator>Chunyuan Lei</creator><creator>Zhu, Xu</creator><creator>Nie, Linzhen</creator><creator>Tao, Bo</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope></search><sort><creationdate>20230324</creationdate><title>A Hierarchical Hybrid Learning Framework for Multi-agent Trajectory Prediction</title><author>Jiao, Yujun ; Miao, Mingze ; Yin, Zhishuai ; Chunyuan Lei ; Zhu, Xu ; Nie, Linzhen ; Tao, Bo</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_27901931813</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Contingency</topic><topic>Deep learning</topic><topic>Kinematics</topic><topic>Motion planning</topic><topic>Multiagent systems</topic><topic>Optimization</topic><topic>Predictions</topic><topic>Trajectories</topic><topic>Transformers</topic><toplevel>online_resources</toplevel><creatorcontrib>Jiao, Yujun</creatorcontrib><creatorcontrib>Miao, Mingze</creatorcontrib><creatorcontrib>Yin, Zhishuai</creatorcontrib><creatorcontrib>Chunyuan Lei</creatorcontrib><creatorcontrib>Zhu, Xu</creatorcontrib><creatorcontrib>Nie, Linzhen</creatorcontrib><creatorcontrib>Tao, Bo</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science & Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Jiao, Yujun</au><au>Miao, Mingze</au><au>Yin, Zhishuai</au><au>Chunyuan Lei</au><au>Zhu, Xu</au><au>Nie, Linzhen</au><au>Tao, Bo</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>A Hierarchical Hybrid Learning Framework for Multi-agent Trajectory Prediction</atitle><jtitle>arXiv.org</jtitle><date>2023-03-24</date><risdate>2023</risdate><eissn>2331-8422</eissn><abstract>Accurate and robust trajectory prediction of neighboring agents is critical for autonomous vehicles traversing in complex scenes. Most methods proposed in recent years are deep learning-based due to their strength in encoding complex interactions. However, unplausible predictions are often generated since they rely heavily on past observations and cannot effectively capture the transient and contingency interactions from sparse samples. In this paper, we propose a hierarchical hybrid framework of deep learning (DL) and reinforcement learning (RL) for multi-agent trajectory prediction, to cope with the challenge of predicting motions shaped by multi-scale interactions. In the DL stage, the traffic scene is divided into multiple intermediate-scale heterogenous graphs based on which Transformer-style GNNs are adopted to encode heterogenous interactions at intermediate and global levels. In the RL stage, we divide the traffic scene into local sub-scenes utilizing the key future points predicted in the DL stage. To emulate the motion planning procedure so as to produce trajectory predictions, a Transformer-based Proximal Policy Optimization (PPO) incorporated with a vehicle kinematics model is devised to plan motions under the dominant influence of microscopic interactions. A multi-objective reward is designed to balance between agent-centric accuracy and scene-wise compatibility. Experimental results show that our proposal matches the state-of-the-arts on the Argoverse forecasting benchmark. It's also revealed by the visualized results that the hierarchical learning framework captures the multi-scale interactions and improves the feasibility and compliance of the predicted trajectories.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><oa>free_for_read</oa></addata></record> |
fulltext | fulltext |
identifier | EISSN: 2331-8422 |
ispartof | arXiv.org, 2023-03 |
issn | 2331-8422 |
language | eng |
recordid | cdi_proquest_journals_2790193181 |
source | Free E- Journals |
subjects | Contingency Deep learning Kinematics Motion planning Multiagent systems Optimization Predictions Trajectories Transformers |
title | A Hierarchical Hybrid Learning Framework for Multi-agent Trajectory Prediction |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-29T02%3A08%3A54IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=document&rft.atitle=A%20Hierarchical%20Hybrid%20Learning%20Framework%20for%20Multi-agent%20Trajectory%20Prediction&rft.jtitle=arXiv.org&rft.au=Jiao,%20Yujun&rft.date=2023-03-24&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E2790193181%3C/proquest%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2790193181&rft_id=info:pmid/&rfr_iscdi=true |