Research map / Multi-agent systems and games / Human-AI and goal-conditioned
TraPHic: Trajectory Prediction in Dense and Heterogeneous Traffic Using Weighted Interactions
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| add horizon-based and heterogeneous-based weighted interactions (ablation) | TraPHic-B on TRAF · TRAF | ADE (RMSE, meters): 2.73 -> 0.78; reducing the ADE by 71% | Helped |
| add horizon-based and heterogeneous-based weighted interactions (ablation) | TraPHic-B on TRAF · TRAF | FDE (RMSE, meters): 7.21 -> 2.44; reducing the FDE by 66% | Helped |
| add horizon-based weighted interactions (ablation) | TraPHic-He on TRAF · TRAF | ADE (RMSE, meters): 2.33 -> 0.78 | Helped |
| add heterogeneous-based weighted interactions (ablation) | TraPHic-Ho on TRAF · TRAF | ADE (RMSE, meters): 1.22 -> 0.78 | Helped |
| use horizon-based weighted interactions without heterogeneous-based weighted interactions (ablation) | TraPHic-B on TRAF · TRAF | ADE (RMSE, meters): 2.73 -> 1.22; TraPHic-Ho reduces ADE by 55% over TraPHic-B | Helped |
| use heterogeneous-based weighted interactions without horizon-based weighted interactions (ablation) | TraPHic-B on TRAF · TRAF | ADE (RMSE, meters): 2.73 -> 2.33; TraPHic-He reduces ADE by 15% over TraPHic-B | Helped |
| apply TraPHic rather than CS-LSTM | CS-LSTM on TRAF · TRAF | ADE (RMSE, meters): 1.15 -> 0.78 | Helped |
| apply TraPHic rather than S-GAN | S-GAN on Beijing · Beijing | ADE (RMSE, meters): 4.02 -> 2.16 | Helped |
| apply TraPHic rather than the strongest listed baseline | S-GAN on NGSIM · NGSIM | ADE (RMSE, meters): 5.16 -> 5.63 | Hurt |
Closest papers by meaning
- Density-Adaptive Model Based on Motif Matrix for Multi-Agent Trajectory Prediction (CVPR 2024 · similarity 0.88)
- Trajeglish: Traffic Modeling as Next-Token Prediction (ICLR 2024 · similarity 0.88)
- HPNet: Dynamic Trajectory Forecasting with Historical Prediction Attention (CVPR 2024 · similarity 0.86)
- NEST: A Neuromodulated Small-world Hypergraph Trajectory Prediction Model for Autonomous Driving (AAAI 2025 · similarity 0.86)
- SEPT: Towards Efficient Scene Representation Learning for Motion Prediction (ICLR 2024 · similarity 0.85)
- DeMo: Decoupling Motion Forecasting into Directional Intentions and Dynamic States (NeurIPS 2024 · similarity 0.85)
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.