Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Stochastic Prediction of Multi-Agent Interactions from Partial Observations
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 |
|---|---|---|---|
| Add stochastic latent dynamics to Graph-RNN (ablation) | Graph-RNN · Basketball | normalized ℓ2 distance at t=5: 0.034 -> 0.024 | Helped |
| Add stochastic latent dynamics to Graph-RNN (ablation) | Graph-RNN · Soccer World | normalized ℓ2 distance at t=5: 0.147 -> 0.143 | Helped |
| Add stochastic latent dynamics to Graph-RNN (ablation) | Graph-RNN · Basketball | log-likelihood ratio over random guess: 15.0 -> ≥ 18.4 | Helped |
| Add stochastic latent dynamics to Graph-RNN (ablation) | Graph-RNN · Soccer World | log-likelihood ratio over random guess: 60.9 -> ≥ 61.1 | Helped |
| Use a graph interaction network instead of independent agent RNNs (ablation) | Indep-RNN · Basketball | log-likelihood ratio over random guess: 14.3 -> 15.0 | Helped |
| Use a graph interaction network instead of Social-RNN pooling (ablation) | Social-RNN · Basketball | log-likelihood ratio over random guess: 14.6 -> 15.0 | Helped |
| Use a graph interaction network instead of independent agent RNNs (ablation) | Indep-RNN · Soccer World | log-likelihood ratio over random guess: 48.9 -> 60.9 | Helped |
| Use a graph interaction network instead of Social-RNN pooling (ablation) | Social-RNN · Soccer World | log-likelihood ratio over random guess: 54.6 -> 60.9 | Helped |
| Add stochastic latent dynamics to Graph-RNN (ablation) | Graph-RNN · Basketball, ball hidden | normalized ℓ2 distance: 0.251 -> 0.227 | Helped |
| Add stochastic latent dynamics to Graph-RNN (ablation) | Graph-RNN · Basketball, player hidden | normalized ℓ2 distance: 0.240 -> 0.214 | Helped |
Closest papers by meaning
- Semi-Supervised Generative Models for Multiagent Trajectories (NeurIPS 2022 · similarity 0.88)
- Diverse Generation for Multi-Agent Sports Games (CVPR 2019 · similarity 0.88)
- Multiple Futures Prediction (NeurIPS 2019 · similarity 0.86)
- Learning Long-term Visual Dynamics with Region Proposal Interaction Networks (ICLR 2021 · similarity 0.86)
- Learning Continuous System Dynamics from Irregularly-Sampled Partial Observations (NeurIPS 2020 · similarity 0.86)
- Event2Tracking: Reconstructing Multi-Agent Soccer Trajectories Using Long-Term Multimodal Context (AAAI 2025 · 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.