Research map / Probabilistic and scientific modeling / Molecules and proteins
What Truly Matters in Trajectory Prediction for Autonomous Driving?
Tran Phong, Haoran Wu, Cunjun Yu, Panpan Cai, Sifa Zheng, David Hsu
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 changed | Where | Measured | Effect |
|---|---|---|---|
| account for the dynamics gap (Dynamic ADE vs Static ADE) | ADE-driving correlation · RVO planner | correlation coefficient: 0.77 | Improved |
| account for the dynamics gap (Dynamic ADE vs Static ADE) | ADE-driving correlation · DESPOT planner | correlation coefficient: 1.16 | Improved |
| use multi-modal prediction (minADE) | ADE-driving correlation · RVO planner | correlation coefficient: 0.70 | Improved |
| use multi-modal prediction (minADE) | ADE-driving correlation · DESPOT planner | correlation coefficient: 0.45 | Improved |
| consider only agents that may affect ego-plan (prediction asymmetry) | ADE-driving correlation · RVO planner | correlation coefficient: 0.17 | Improved |
| consider only agents that may affect ego-plan (prediction asymmetry) | ADE-driving correlation · DESPOT planner | correlation coefficient: 0.20 | Improved |
| consider agents with missed observations (occlusion) | ADE-driving correlation · RVO planner | correlation coefficient: -0.22 | Hurt |
| consider agents with missed observations (occlusion) | ADE-driving correlation · DESPOT planner | correlation coefficient: 0.16 | Improved |
| add temporal consistency of predictions as indicator | driving performance · RVO and DESPOT planners | correlation | No clear effect |
| add error variance (error distribution) as indicator | driving performance · RVO and DESPOT planners | correlation | No clear effect |
| use faster predictor CV instead of KNN | DESPOT planner · 30Hz tick | driving performance: 0.526 -> 0.753 | Improved |
| raise tick rate from 1Hz to 30Hz | LSTM · DESPOT planner | driving performance: 0.766 -> 0.644 | Hurt |
Closest papers by meaning
- Evaluation of Trajectory Distribution Predictions with Energy Score (ICML 2024 · similarity 0.86)
- BAT: Behavior-Aware Human-Like Trajectory Prediction for Autonomous Driving (AAAI 2024 · similarity 0.85)
- Certified Human Trajectory Prediction (CVPR 2025 · similarity 0.85)
- CCTR: Calibrating Trajectory Prediction for Uncertainty-Aware Motion Planning in Autonomous Driving (AAAI 2024 · similarity 0.84)
- LaKD: Length-agnostic Knowledge Distillation for Trajectory Prediction with Any Length Observations (NeurIPS 2024 · similarity 0.84)
- Leveraging Driver Field-of-View for Multimodal Ego-Trajectory Prediction (ICLR 2025 · similarity 0.83)
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.
