Research map / Graph learning / Knowledge and temporal graphs
ViTE: Virtual Graph Trajectory Expert Router for Pedestrian Trajectory Prediction
Ruochen Li, Zhanxing Zhu, Tanqiu Qiao, Hubert P. H. Shum
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 adaptive expert router versus no expert (ablation) | ViTE · ETH/UCY | min ADE20: 0.25 -> 0.20 | Improved |
| add adaptive expert router versus no expert (ablation) | ViTE · ETH/UCY | min FDE20: 0.38 -> 0.32 | Improved |
| use adaptive MoE routing instead of fixed mixed experts (ablation) | ViTE · ETH/UCY | min ADE20: 0.25 -> 0.20 | Improved |
| fixed mixed expert versus no expert (ablation) | ViTE · ETH/UCY | min ADE20: 0.25 -> 0.25 | No clear effect |
| use high-order expert only instead of one-hop expert only (ablation) | ViTE · ETH/UCY | min ADE20: 0.22 -> 0.23 | Hurt |
| use virtual graph instead of 2-layer GCN (ablation) | ViTE · ETH/UCY | min ADE20: 0.22 -> 0.20 | Improved |
| use virtual graph instead of 4-layer GCN (ablation) | ViTE · ETH/UCY | min FDE20: 0.36 -> 0.32 | Improved |
| increase GCN depth from 2 to 4 layers (ablation) | GCN baseline in ViTE · ETH/UCY | min ADE20: 0.22 -> 0.23 | Hurt |
| ViTE vs PCHGCN on UNIV | ViTE · ETH/UCY UNIV | min ADE20: 0.21 -> 0.23 | Hurt |
| ViTE vs PCHGCN on average | ViTE · ETH/UCY average | min FDE20: 0.37 -> 0.32 | Improved |
| ViTE vs MART on NBA 4.0s | ViTE · NBA 4.0s | min FDE20: 0.90 -> 0.91 | Hurt |
| ViTE vs LED on SDD | ViTE · SDD 4.8s | min FDE20: 11.66 -> 11.90 | Hurt |
Closest papers by meaning
- A Set of Control Points Conditioned Pedestrian Trajectory Prediction (AAAI 2023 · similarity 0.89)
- SGCN: Sparse Graph Convolution Network for Pedestrian Trajectory Prediction (CVPR 2021 · similarity 0.89)
- Higher-order Relational Reasoning for Pedestrian Trajectory Prediction (CVPR 2024 · similarity 0.88)
- A Unified Environmental Network for Pedestrian Trajectory Prediction (AAAI 2024 · similarity 0.87)
- MS-TIP: Imputation Aware Pedestrian Trajectory Prediction (ICML 2024 · similarity 0.86)
- SocialCVAE: Predicting Pedestrian Trajectory via Interaction Conditioned Latents (AAAI 2024 · 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.
