Research map / Multi-agent systems and games / Human-AI and goal-conditioned
MUSE-VAE: Multi-Scale VAE for Environment-Aware Long Term Trajectory Prediction
Mihee Lee, Samuel S. Sohn, Seonghyeon Moon, Sejong Yoon, Mubbasir Kapadia, Vladimir Pavlovic
Check 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 |
|---|---|---|---|
| Use MUSE-VAE instead of the strongest baseline | PFSD (K=50) · PFSD | KDE NLL: 1.17 -> -1.94 | Improved |
| Use MUSE-VAE instead of the strongest baseline | PFSD (K=50) · PFSD | ADE: 0.08 -> 0.07 | Mixed |
| Use MUSE-VAE instead of the strongest baseline | SDD (K=20) · SDD | ADE: 7.84 -> 6.36 | Improved |
| Use MUSE-VAE instead of the strongest baseline | SDD (K=20) · SDD | ECFL: 89.32 -> 89.30 | No clear effect |
| Use MUSE-VAE instead of the strongest baseline | nuScenes (K=10) · nuScenes | ADE: 1.30 -> 1.09 | Improved |
| Add SG-net waypoint predictions (ablation) | MUSE-VAE · PFSD (K=20) | KDE NLL: -1.39 -> -1.66 | Improved |
| Add SG-net waypoint predictions (ablation) | MUSE-VAE · PFSD (K=20) | ADE: 0.10 -> 0.09 | No clear effect |
| Add Micro-stage trajectory refinement (ablation) | MUSE-VAE · PFSD (K=20) | ECFL: 99.75 -> 97.40 | Hurt |
| Add Micro-stage trajectory refinement (ablation) | MUSE-VAE · PFSD (K=20) | ADE: 0.10 -> 0.09 | Improved |
| Add prior-distribution reconstruction loss (ablation) | MUSE-VAE · PFSD (K=20) | ECFL: 95.06 -> 97.40 | Improved |
| Add prior-distribution reconstruction loss (ablation) | MUSE-VAE · PFSD (K=20) | KDE NLL: -1.77 -> -1.66 | Hurt |
Later papers that compared with this method
1 later accepted paper reports MUSE-VAE as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Density-Adaptive Model Based on Motif Matrix for Multi-Agent Trajectory Prediction (CVPR 2024 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- AgentFormer — AgentFormer: Agent-Aware Transformers for Socio-Temporal Multi-Agent Forecasting (ICCV 2021) · better on 5 of 7 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
- Multiple Futures Prediction (NeurIPS 2019 · similarity 0.88)
- Multi-agent Long-term 3D Human Pose Forecasting via Interaction-aware Trajectory Conditioning (CVPR 2024 · similarity 0.87)
- HPNet: Dynamic Trajectory Forecasting with Historical Prediction Attention (CVPR 2024 · similarity 0.86)
- AutoTraces: Autoregressive Trajectory Forecasting via Multimodal Large Language Models (CVPR 2026 · similarity 0.86)
- Multi-modal Knowledge Distillation-based Human Trajectory Forecasting (CVPR 2025 · 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.
