Research map / Multi-agent systems and games / Human-AI and goal-conditioned
The Trajectron: Probabilistic Multi-Agent Trajectory Modeling With Dynamic Spatiotemporal Graphs
Boris Ivanovic, Marco Pavone
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 |
|---|---|---|---|
| sample the full latent z distribution instead of fixing z to its mode (ablation) | Trajectron · ETH/UCY | mean NLL | Improved |
| use Trajectron (Full) instead of SGAN | Trajectron · ETH/UCY average (best of 100) | ADE (m): 0.43 -> 0.34 | Improved |
| use Trajectron (Full) instead of SGAN | Trajectron · ETH/UCY average (best of 100) | FDE (m): 0.86 -> 0.67 | Improved |
| use Trajectron (Full) instead of SGAN | Trajectron · Zara 1 (best of 100) | ADE (m): 0.29 -> 0.32 | Hurt |
| use Trajectron (Full) instead of SGAN | Trajectron · Zara 2 (best of 100) | ADE (m): 0.27 -> 0.34 | Hurt |
| use Trajectron (Full) instead of SGAN | Trajectron · ETH (best of 100) | ADE (m): 0.64 -> 0.37 | Improved |
| use Trajectron (Full) instead of SGAN | Trajectron · Zara 1 | NLL | Hurt |
| use Trajectron instead of SGAN | Trajectron · ETH, Hotel, Univ and average | NLL | Improved |
| use Trajectron (Full) instead of SGAN | Trajectron · Zara 2 | NLL | No clear effect |
| use Trajectron instead of SGAN | Trajectron · ETH | runtime for 200 samples (s): 6.98 -> 0.13 | Improved |
| use z_best instead of Full sampling (ablation) | Trajectron · Univ | runtime for 200 samples (s): 2.00 -> 1.96 | No clear effect |
Closest papers by meaning
- Trajectron++: Dynamically-Feasible Trajectory Forecasting With Heterogeneous Data (ECCV 2020 · similarity 0.94)
- AgentFormer: Agent-Aware Transformers for Socio-Temporal Multi-Agent Forecasting (ICCV 2021 · similarity 0.89)
- Scene Transformer: A unified architecture for predicting future trajectories of multiple agents (ICLR 2022 · similarity 0.89)
- SMART: Simultaneous Multi-Agent Recurrent Trajectory Prediction (ECCV 2020 · similarity 0.89)
- AutoTraces: Autoregressive Trajectory Forecasting via Multimodal Large Language Models (CVPR 2026 · similarity 0.87)
- SmartRefine: A Scenario-Adaptive Refinement Framework for Efficient Motion Prediction (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.
