Research map / Probabilistic and scientific modeling / Molecules and proteins
NATRA: Noise-Agnostic Framework for Trajectory Prediction with Noisy Observations
Rongqing Li, Changsheng Li, Ruilin Lv, Yuhang Li, Yang Gao, Xiaolu Zhang, Jun Zhou
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 NATRA denoising | GraphTern · ETH/UCY avg (σ=0.4) | ADE: 0.51 -> 0.40 | Improved |
| add NATRA denoising | EqMotion · ETH/UCY avg (σ=0.4) | ADE: 0.42 -> 0.35 | Improved |
| add NATRA denoising | EqMotion · SDD (σ=0.4) | ADE: 13.46 -> 11.92 | Improved |
| add NATRA denoising | HiVT · Argoverse K=1 (σ=4) | ADE: 3.16 -> 2.78 | Improved |
| add NATRA denoising | GraphTern · ETH/UCY noisy subset | ADE: 0.48 -> 0.42 | Improved |
| add NATRA denoising | EqMotion · SDD (Poisson noise λ=0.4) | ADE: 14.05 -> 12.22 | Improved |
| add NATRA denoising | EqMotion · SDD (mixed Gaussian+Poisson noise) | ADE: 14.66 -> 12.96 | Improved |
| use NATRA instead of Wavelet denoising | EqMotion · SDD (σ=0.4) | ADE: 12.38 -> 11.92 | Improved |
| use NATRA instead of EMA denoising | GraphTern · ETH/UCY avg (σ=0.4) | ADE: 0.47 -> 0.40 | Improved |
| add max term of mutual information between denoised and clean trajectories (ablation) | NATRA (GraphTern, σ=0.4) · ETH/UCY | ADE: 0.51 -> 0.48 | Improved |
| add min term of mutual information between denoised and noisy trajectories (ablation) | NATRA (GraphTern, σ=0.4) · ETH/UCY | ADE: 0.48 -> 0.47 | Improved |
| add reconstruction loss Lrec (ablation) | NATRA (GraphTern, σ=0.4) · ETH/UCY | ADE: 0.47 -> 0.42 | Improved |
Closest papers by meaning
- C2F-TP: A Coarse-to-Fine Denoising Framework for Uncertainty-Aware Trajectory Prediction (AAAI 2025 · similarity 0.88)
- Certified Human Trajectory Prediction (CVPR 2025 · similarity 0.87)
- Evaluation of Trajectory Distribution Predictions with Energy Score (ICML 2024 · similarity 0.85)
- Proper Scoring Rules for Agentic Uncertainty Quantification (NeurIPS 2026 · similarity 0.85)
- Improving Transferability for Cross-Domain Trajectory Prediction via Neural Stochastic Differential Equation (AAAI 2024 · similarity 0.84)
- SmartRefine: A Scenario-Adaptive Refinement Framework for Efficient Motion Prediction (CVPR 2024 · 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.
