Research map / Probabilistic and scientific modeling / Molecules and proteins
Intention-Aware Diffusion Model for Pedestrian Trajectory Prediction
Yu Liu, Zhijie Liu, Xiao Ren, Youfu Li, He Kong
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 |
|---|---|---|---|
| set number of endpoint candidates M to 5 rather than 1 (ablation) | IAD · ETH/UCY (average) | ADE: 0.24 -> 0.19 | Improved |
| increase endpoint candidates M from 5 to 20 (ablation) | IAD · ETH/UCY (average) | ADE: 0.19 -> 0.27 | Hurt |
| set diffusion steps K to 100 rather than 10 (ablation) | IAD · ETH/UCY (average) | ADE: 0.22 -> 0.19 | Improved |
| increase diffusion steps from 100 to 200 (ablation) | IAD · ETH/UCY (average) | ADE: 0.19 -> 0.23 | Hurt |
| add short-term residual polar intention branch (ablation) | IAD · ETH/UCY and SDD | ADE/FDE | Improved |
| add long-term endpoint intention branch (ablation) | IAD · ETH/UCY and SDD | ADE/FDE | Improved |
| add soft-mask guidance and residual noise refinement (ablation) | IAD · ETH/UCY and SDD | ADE/FDE | Improved |
| IAD versus PPT on SDD ADE | SDD · SDD | ADE: 7.03 -> 6.85 | Improved |
| IAD versus MoFlow on NBA total horizon | NBA · NBA (4.0s total) | ADE: 0.71 -> 0.79 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- MoFlow — MoFlow: One-Step Flow Matching for Human Trajectory Forecasting via Implicit Maximum Likelihood Estimation based Distillation (CVPR 2025) · better on 3 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Condition-Guided Diffusion for Multi-Modal Pedestrian Trajectory Prediction Incorporating Intention and Interaction Priors (TPAMI 2025 · similarity 0.91)
- BCDiff: Bidirectional Consistent Diffusion for Instantaneous Trajectory Prediction (NeurIPS 2023 · similarity 0.91)
- Leapfrog Diffusion Model for Stochastic Trajectory Prediction (CVPR 2023 · similarity 0.90)
- SocialMOIF: Multi-Order Intention Fusion for Pedestrian Trajectory Prediction (CVPR 2025 · similarity 0.88)
- Towards Predicting Any Human Trajectory In Context (NeurIPS 2025 · similarity 0.88)
- SingularTrajectory: Universal Trajectory Predictor Using Diffusion Model (CVPR 2024 · similarity 0.87)
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.
