Research map / Probabilistic and scientific modeling / Molecules and proteins
AMD: Adaptive Momentum and Decoupled Contrastive Learning Framework for Robust Long-Tail Trajectory Prediction
Bin Rao, Haicheng Liao, Yanchen Guan, Chengyue Wang, Bonan Wang, Jiaxun Zhang, Zhenning Li
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 |
|---|---|---|---|
| add trajectory augmentation (ablation) | AMD · nuScenes Top 1% | minFDE: 2.05 -> 1.66 | Improved |
| add MoCo-DT (ablation) | AMD · nuScenes Top 1% | minFDE: 1.79 -> 1.66 | Improved |
| add online iterative clustering (ablation) | AMD · nuScenes Top 1% | minFDE: 1.95 -> 1.66 | Improved |
| add decoupled contrastive learning (ablation) | AMD · nuScenes Top 1% | minFDE: 2.02 -> 1.66 | Improved |
| add all four components (ablation) | AMD · nuScenes Top 1% | minFDE: 2.41 -> 1.66 | Improved |
| loss weights lambda1,lambda2 from 0.5,0.1 to 1.0,0.1 (ablation) | AMD · nuScenes Top 1% | minFDE: 2.00 -> 1.66 | Improved |
| momentum coefficients 0.95,0.99,0.999 vs 0.90,0.95,0.999 (ablation) | AMD · nuScenes Top 1% | minFDE: 2.04 -> 1.66 | Improved |
| AMD vs FEND on long tail | AMD · nuScenes Top 1% | minFDE: 2.50 -> 1.66 | Improved |
| AMD vs FEND on all samples (2s) | AMD · nuScenes All | minADE: 0.17 -> 0.21 | Hurt |
| AMD vs PPT on ETH/UCY average | AMD · ETH/UCY average | minFDE: 0.31 -> 0.27 | Improved |
| AMD vs LaPred miss rate | AMD · nuScenes (6s) | MR_5: 0.53 -> 0.50 | Improved |
| AMD vs SeFlow on minADE_10 | AMD · nuScenes (6s) | minADE_10: 0.98 -> 1.06 | Hurt |
Closest papers by meaning
- C2F-TP: A Coarse-to-Fine Denoising Framework for Uncertainty-Aware Trajectory Prediction (AAAI 2025 · similarity 0.85)
- Adapting to Observation Length of Trajectory Prediction via Contrastive Learning (CVPR 2025 · similarity 0.85)
- LaKD: Length-agnostic Knowledge Distillation for Trajectory Prediction with Any Length Observations (NeurIPS 2024 · similarity 0.85)
- STraj: Self-training for Bridging the Cross-Geography Gap in Trajectory Prediction (AAAI 2025 · similarity 0.85)
- DeMo: Decoupling Motion Forecasting into Directional Intentions and Dynamic States (NeurIPS 2024 · similarity 0.85)
- Don't Shake the Wheel: Momentum-Aware Planning in End-to-End Autonomous Driving (CVPR 2025 · similarity 0.84)
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.
