Divide-and-Conquer for Lane-Aware Diverse Trajectory Prediction

Trajectory prediction is a safety-critical tool for autonomous vehicles to plan and execute actions. Our work addresses two key challenges in trajectory prediction, learning multimodal outputs, and better predictions by imposing constraints using driving knowledge. Recent methods have achieved stron...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Narayanan, Sriram, Moslemi, Ramin, Pittaluga, Francesco, Liu, Buyu, Chandraker, Manmohan
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Narayanan, Sriram
Moslemi, Ramin
Pittaluga, Francesco
Liu, Buyu
Chandraker, Manmohan
description Trajectory prediction is a safety-critical tool for autonomous vehicles to plan and execute actions. Our work addresses two key challenges in trajectory prediction, learning multimodal outputs, and better predictions by imposing constraints using driving knowledge. Recent methods have achieved strong performances using Multi-Choice Learning objectives like winner-takes-all (WTA) or best-of-many. But the impact of those methods in learning diverse hypotheses is under-studied as such objectives highly depend on their initialization for diversity. As our first contribution, we propose a novel Divide-And-Conquer (DAC) approach that acts as a better initialization technique to WTA objective, resulting in diverse outputs without any spurious modes. Our second contribution is a novel trajectory prediction framework called ALAN that uses existing lane centerlines as anchors to provide trajectories constrained to the input lanes. Our framework provides multi-agent trajectory outputs in a forward pass by capturing interactions through hypercolumn descriptors and incorporating scene information in the form of rasterized images and per-agent lane anchors. Experiments on synthetic and real data show that the proposed DAC captures the data distribution better compare to other WTA family of objectives. Further, we show that our ALAN approach provides on par or better performance with SOTA methods evaluated on Nuscenes urban driving benchmark.
doi_str_mv 10.48550/arxiv.2104.08277
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2104_08277</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2104_08277</sourcerecordid><originalsourceid>FETCH-LOGICAL-a677-89df873ea7d4b14e753fa2c34239771dbd2db6ea2b5d3cff65b8bc3909ca2ebb3</originalsourceid><addsrcrecordid>eNotz7tOwzAUxnEvDKjlAZjwCzg4vsTOVFXhKkWCIXt0bB9LRpDAaenl7YHS6Vv--qQfY9e1rIy3Vt4CHcquUrU0lfTKuUu2uiu7klDAlEQ3T1_fSDzPxHuYUKz3QMh_C6QN8oHgDeN2piN_JUwlbss8LdlFhvcNXp13wYaH-6F7Ev3L43O37gU0zgnfpuydRnDJhNqgszqDitoo3TpXp5BUCg2CCjbpmHNjgw9Rt7KNoDAEvWA3_7cnwfhJ5QPoOP5JxpNE_wC_7kQ6</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Divide-and-Conquer for Lane-Aware Diverse Trajectory Prediction</title><source>arXiv.org</source><creator>Narayanan, Sriram ; Moslemi, Ramin ; Pittaluga, Francesco ; Liu, Buyu ; Chandraker, Manmohan</creator><creatorcontrib>Narayanan, Sriram ; Moslemi, Ramin ; Pittaluga, Francesco ; Liu, Buyu ; Chandraker, Manmohan</creatorcontrib><description>Trajectory prediction is a safety-critical tool for autonomous vehicles to plan and execute actions. Our work addresses two key challenges in trajectory prediction, learning multimodal outputs, and better predictions by imposing constraints using driving knowledge. Recent methods have achieved strong performances using Multi-Choice Learning objectives like winner-takes-all (WTA) or best-of-many. But the impact of those methods in learning diverse hypotheses is under-studied as such objectives highly depend on their initialization for diversity. As our first contribution, we propose a novel Divide-And-Conquer (DAC) approach that acts as a better initialization technique to WTA objective, resulting in diverse outputs without any spurious modes. Our second contribution is a novel trajectory prediction framework called ALAN that uses existing lane centerlines as anchors to provide trajectories constrained to the input lanes. Our framework provides multi-agent trajectory outputs in a forward pass by capturing interactions through hypercolumn descriptors and incorporating scene information in the form of rasterized images and per-agent lane anchors. Experiments on synthetic and real data show that the proposed DAC captures the data distribution better compare to other WTA family of objectives. Further, we show that our ALAN approach provides on par or better performance with SOTA methods evaluated on Nuscenes urban driving benchmark.</description><identifier>DOI: 10.48550/arxiv.2104.08277</identifier><language>eng</language><subject>Computer Science - Computer Vision and Pattern Recognition</subject><creationdate>2021-04</creationdate><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,881</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2104.08277$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2104.08277$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Narayanan, Sriram</creatorcontrib><creatorcontrib>Moslemi, Ramin</creatorcontrib><creatorcontrib>Pittaluga, Francesco</creatorcontrib><creatorcontrib>Liu, Buyu</creatorcontrib><creatorcontrib>Chandraker, Manmohan</creatorcontrib><title>Divide-and-Conquer for Lane-Aware Diverse Trajectory Prediction</title><description>Trajectory prediction is a safety-critical tool for autonomous vehicles to plan and execute actions. Our work addresses two key challenges in trajectory prediction, learning multimodal outputs, and better predictions by imposing constraints using driving knowledge. Recent methods have achieved strong performances using Multi-Choice Learning objectives like winner-takes-all (WTA) or best-of-many. But the impact of those methods in learning diverse hypotheses is under-studied as such objectives highly depend on their initialization for diversity. As our first contribution, we propose a novel Divide-And-Conquer (DAC) approach that acts as a better initialization technique to WTA objective, resulting in diverse outputs without any spurious modes. Our second contribution is a novel trajectory prediction framework called ALAN that uses existing lane centerlines as anchors to provide trajectories constrained to the input lanes. Our framework provides multi-agent trajectory outputs in a forward pass by capturing interactions through hypercolumn descriptors and incorporating scene information in the form of rasterized images and per-agent lane anchors. Experiments on synthetic and real data show that the proposed DAC captures the data distribution better compare to other WTA family of objectives. Further, we show that our ALAN approach provides on par or better performance with SOTA methods evaluated on Nuscenes urban driving benchmark.</description><subject>Computer Science - Computer Vision and Pattern Recognition</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotz7tOwzAUxnEvDKjlAZjwCzg4vsTOVFXhKkWCIXt0bB9LRpDAaenl7YHS6Vv--qQfY9e1rIy3Vt4CHcquUrU0lfTKuUu2uiu7klDAlEQ3T1_fSDzPxHuYUKz3QMh_C6QN8oHgDeN2piN_JUwlbss8LdlFhvcNXp13wYaH-6F7Ev3L43O37gU0zgnfpuydRnDJhNqgszqDitoo3TpXp5BUCg2CCjbpmHNjgw9Rt7KNoDAEvWA3_7cnwfhJ5QPoOP5JxpNE_wC_7kQ6</recordid><startdate>20210416</startdate><enddate>20210416</enddate><creator>Narayanan, Sriram</creator><creator>Moslemi, Ramin</creator><creator>Pittaluga, Francesco</creator><creator>Liu, Buyu</creator><creator>Chandraker, Manmohan</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20210416</creationdate><title>Divide-and-Conquer for Lane-Aware Diverse Trajectory Prediction</title><author>Narayanan, Sriram ; Moslemi, Ramin ; Pittaluga, Francesco ; Liu, Buyu ; Chandraker, Manmohan</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a677-89df873ea7d4b14e753fa2c34239771dbd2db6ea2b5d3cff65b8bc3909ca2ebb3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Computer Science - Computer Vision and Pattern Recognition</topic><toplevel>online_resources</toplevel><creatorcontrib>Narayanan, Sriram</creatorcontrib><creatorcontrib>Moslemi, Ramin</creatorcontrib><creatorcontrib>Pittaluga, Francesco</creatorcontrib><creatorcontrib>Liu, Buyu</creatorcontrib><creatorcontrib>Chandraker, Manmohan</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Narayanan, Sriram</au><au>Moslemi, Ramin</au><au>Pittaluga, Francesco</au><au>Liu, Buyu</au><au>Chandraker, Manmohan</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Divide-and-Conquer for Lane-Aware Diverse Trajectory Prediction</atitle><date>2021-04-16</date><risdate>2021</risdate><abstract>Trajectory prediction is a safety-critical tool for autonomous vehicles to plan and execute actions. Our work addresses two key challenges in trajectory prediction, learning multimodal outputs, and better predictions by imposing constraints using driving knowledge. Recent methods have achieved strong performances using Multi-Choice Learning objectives like winner-takes-all (WTA) or best-of-many. But the impact of those methods in learning diverse hypotheses is under-studied as such objectives highly depend on their initialization for diversity. As our first contribution, we propose a novel Divide-And-Conquer (DAC) approach that acts as a better initialization technique to WTA objective, resulting in diverse outputs without any spurious modes. Our second contribution is a novel trajectory prediction framework called ALAN that uses existing lane centerlines as anchors to provide trajectories constrained to the input lanes. Our framework provides multi-agent trajectory outputs in a forward pass by capturing interactions through hypercolumn descriptors and incorporating scene information in the form of rasterized images and per-agent lane anchors. Experiments on synthetic and real data show that the proposed DAC captures the data distribution better compare to other WTA family of objectives. Further, we show that our ALAN approach provides on par or better performance with SOTA methods evaluated on Nuscenes urban driving benchmark.</abstract><doi>10.48550/arxiv.2104.08277</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2104.08277
ispartof
issn
language eng
recordid cdi_arxiv_primary_2104_08277
source arXiv.org
subjects Computer Science - Computer Vision and Pattern Recognition
title Divide-and-Conquer for Lane-Aware Diverse Trajectory Prediction
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-06T03%3A02%3A14IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Divide-and-Conquer%20for%20Lane-Aware%20Diverse%20Trajectory%20Prediction&rft.au=Narayanan,%20Sriram&rft.date=2021-04-16&rft_id=info:doi/10.48550/arxiv.2104.08277&rft_dat=%3Carxiv_GOX%3E2104_08277%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true