Research map / Probabilistic and scientific modeling / Molecules and proteins
Stochastic Trajectory Prediction via Motion Indeterminacy Diffusion
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 |
|---|---|---|---|
| Use motion indeterminacy diffusion for trajectory generation (ablation) | MID versus Trajectron++ with the LSTM decoder · Stanford Drone Dataset | ADE: 8.98 -> 7.61 | Helped |
| Use motion indeterminacy diffusion for trajectory generation (ablation) | MID versus Trajectron++ with the LSTM decoder · Stanford Drone Dataset | FDE: 19.02 -> 14.30 | Helped |
| Replace the LSTM with a Transformer decoder in the CVAE-based Trajectron++ framework (ablation) | Trajectron++ · Stanford Drone Dataset | ADE: 8.98 -> 9.86 | Hurt |
| Replace the LSTM with a Transformer decoder in the CVAE-based Trajectron++ framework (ablation) | Trajectron++ · Stanford Drone Dataset | FDE: 19.02 -> 19.56 | Hurt |
| Use a Transformer decoder instead of a linear decoder (ablation) | MID · Stanford Drone Dataset | ADE: 8.85 -> 7.61 | Helped |
| Use a Transformer decoder instead of an LSTM decoder (ablation) | MID · Stanford Drone Dataset | ADE: 8.41 -> 7.61 | Helped |
| Increase Transformer width from 512d to 1024d (ablation) | MID · Stanford Drone Dataset | ADE: 7.61 -> 7.64 | Hurt |
| Increase Transformer width from 512d to 1024d (ablation) | MID · Stanford Drone Dataset | FDE: 14.30 -> 14.37 | Hurt |
| Reduce Transformer width from 512d to 256d (ablation) | MID · Stanford Drone Dataset | ADE: 7.61 -> 7.91 | Hurt |
| Increase the number of samples from 20 to 40 (ablation) | MID · Stanford Drone Dataset | ADE: 7.61 -> 6.84 | Helped |
| Increase the number of samples from 20 to 20×20 (ablation) | MID · Stanford Drone Dataset | ADE: 7.61 -> 5.42 | Helped |
| Use MID with 20 samples instead of Y-Net + TTST with 10000 samples | MID versus Y-Net + TTST · Stanford Drone Dataset | ADE: 7.85 -> 7.61 | Helped |
Closest papers by meaning
- SocialCVAE: Predicting Pedestrian Trajectory via Interaction Conditioned Latents (AAAI 2024 · similarity 0.88)
- SingularTrajectory: Universal Trajectory Predictor Using Diffusion Model (CVPR 2024 · similarity 0.88)
- Social-Transmotion: Promptable Human Trajectory Prediction (ICLR 2024 · similarity 0.87)
- MS-TIP: Imputation Aware Pedestrian Trajectory Prediction (ICML 2024 · similarity 0.87)
- Pedestrian Trajectory Prediction with Missing Data: Datasets, Imputation, and Benchmarking (NeurIPS 2024 · similarity 0.86)
- Towards Predicting Any Human Trajectory In Context (NeurIPS 2025 · similarity 0.86)
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.