Research map / Probabilistic and scientific modeling / Molecules and proteins
Unsupervised Sampling Promoting for Stochastic Human Trajectory Prediction
Read the paperCheck 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 |
|---|---|---|---|
| Replace MC sampling with sequential Bayesian-optimization sampling | Stochastic trajectory predictors · ETH-UCY full | minADE: 0.53 -> 0.52 (Social-GAN); 0.21 -> 0.18 (Trajectron++); 0.32 -> 0.30 (PECNet); 0.45 -> 0.41 (Social-STGCNN); 0.46 -> 0.4 | Helped |
| Replace MC sampling with sequential Bayesian-optimization sampling | Stochastic trajectory predictors · ETH-UCY exception subset | ADE: 0.94 -> 0.78 (Social-GAN); 0.30 -> 0.27 (Trajectron++); 1.21 -> 0.97 (PECNet); 0.82 -> 0.52 (Social-STGCNN); 0.97 -> 0.7 | Helped |
| Use BOSampler instead of QMC sampling | Stochastic trajectory predictors · ETH-UCY full | minADE: 0.53 -> 0.52 (Social-GAN); 0.21 -> 0.18 (Trajectron++); 0.31 -> 0.30 (PECNet); 0.39 -> 0.41 (Social-STGCNN); 0.45 -> 0.4 | Mixed |
| Add QMC sampling in the BOSampler warm-up stage (ablation) | BOSampler · ETH-UCY full, Social-STGCNN | minADE: 0.41 -> 0.37 | Helped |
| Add QMC sampling in the BOSampler warm-up stage (ablation) | BOSampler · ETH-UCY full, Social-STGCNN | minFDE: 0.69 -> 0.62 | Helped |
| Set warm-up count near half of the total sample count (ablation) | BOSampler with PECNet · ETH-UCY full | ADE and FDE | Helped |
| Reduce the warm-up count to increase BOSampler exploration (ablation) | BOSampler with PECNet · ETH-UCY full | ADE and FDE | Hurt |
| Vary acquisition-function beta within [0.1,1] (ablation) | BOSampler with PECNet · ETH-UCY full | ADE and FDE | No clear effect |
| Increase the number of generated samples (ablation) | BOSampler · ETH-UCY full | ADE and FDE | Helped |
Closest papers by meaning
- SingularTrajectory: Universal Trajectory Predictor Using Diffusion Model (CVPR 2024 · similarity 0.87)
- ProDiff: Prototype-Guided Diffusion for Minimal Information Trajectory Imputation (ICML 2025 · similarity 0.86)
- Physical Plausibility-aware Trajectory Prediction via Locomotion Embodiment (CVPR 2025 · similarity 0.85)
- Stop the Sampler! Classifier-Based Adaptive Stopping for Sampling Kernels (NeurIPS 2026 · similarity 0.85)
- Existence Is Chaos: Enhancing 3D Human Motion Prediction with Uncertainty Consideration (AAAI 2024 · similarity 0.85)
- Model-based Diffusion for Trajectory Optimization (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.