AI Paper Atlas

Research map / Graph learning / Knowledge and temporal graphs

HDGT: Heterogeneous Driving Graph Transformer for Multi-Agent Trajectory Prediction via Scene Encoding

Xiaosong Jia, Penghao Wu, Li Chen, Yu Liu, Hongyang Li, Junchi Yan

TPAMI 2023 · Graph learning › Knowledge and temporal graphs · approach: Heterogeneous graph learning · 237 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add relative (node-local) coordinate encoding (ablation)HDGT · INTERACTION valminADE: 0.2287 -> 0.1889Improved
add HD-Map information (ablation)HDGT · INTERACTION valminADE: 0.1889 -> 0.1526Improved
add semantic-relation graph instead of fully connected graph (ablation)HDGT · INTERACTION valminADE: 0.1526 -> 0.1328Improved
add type-specific heterogeneous parameters (ablation)HDGT · INTERACTION valminADE: 0.1328 -> 0.1071Improved
use Pose Change encoding instead of Fixed Reference (ablation)HDGT · INTERACTION valminADE: 0.1549 -> 0.1081Improved
use Pose Change encoding instead of Relative Edge (ablation)HDGT · INTERACTION valminADE: 0.1158 -> 0.1081Improved
separate lane connectivity edge types instead of merging (ablation)HDGT · INTERACTION valminADE: 0.1622 -> 0.1071Improved
separate lane and traffic sign node types (ablation)HDGT · INTERACTION valminADE: 0.1272 -> 0.1071Improved
use Transformer aggregation instead of GCN-like function (ablation)HDGT · INTERACTION valminADE: 0.1882 -> 0.1071Improved
increase layers from 3 to 6HDGT · INTERACTION valminADE: 0.1071 -> 0.1082No clear effect
increase layers from 1 to 6HDGT · Waymo valminADE: 0.6462 -> 0.5673Improved
increase layers from 3 to 6HDGT · Waymo valinference time: 1038 -> 1146Hurt

Later papers that compared with this method

1 later accepted paper reports HDGT as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy