Research map / Multi-agent systems and games / Human-AI and goal-conditioned
PRECOG: PREdiction Conditioned on Goals in Visual Multi-Agent Settings
Nicholas Rhinehart, Rowan McAllister, Kris Kitani, Sergey Levine
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 LIDAR visual context input (ablation) | ESP · CARLA Town02 Test, 2 agents | minMSD (K=12): 0.632597 -> 0.3925 | Improved |
| add LIDAR visual context input (ablation) | ESP · CARLA Town02 Test, 5 agents | minMSD (K=12): 0.783829 -> 0.564709 | Improved |
| add LIDAR visual context input (ablation) | ESP · nuScenes Test, 2 agents | minMSD (K=12): 1.4964 -> 1.3248 | Improved |
| add LIDAR visual context input (ablation) | ESP · CARLA Town01 Test, 1 agent | minMSD (K=12): 0.49561 -> 0.13589 | Improved |
| add LIDAR visual context input (ablation) | ESP · nuScenes Test, 2 agents | extra nats (e-hat): 0.91982 -> 0.93349 | Hurt |
| add binary road mask to LIDAR input (ablation) | ESP · nuScenes Test, 2 agents | minMSD (K=12): 1.3248 -> 1.0809 | Improved |
| add binary road mask to LIDAR input (ablation) | ESP · nuScenes Test, 5 agents | minMSD (K=12): 3.266 -> 2.8924 | Improved |
| model agents jointly with ESP instead of independently (R2P2-MA) | R2P2-MA · CARLA Town02 Test, 2 agents | minMSD (K=12): 0.454215 -> 0.3925 | Improved |
| model agents jointly with ESP instead of independently (R2P2-MA) | R2P2-MA · nuScenes Test, 5 agents | minMSD (K=12): 3.3109 -> 2.8924 | Improved |
| use flexible-count ESP instead of fixed-count ESP | ESP · CARLA Town02 Test, 5 agents | minMSD (K=12): 0.564709 -> 0.435110 | Improved |
| use flexible-count ESP instead of fixed-count ESP | ESP · CARLA Town02 Test, 2 agents | minMSD (K=12): 0.3925 -> 0.488494 | Hurt |
| train with noise injection (epsilon 0.1) against localization error (ablation) | ESP · nuScenes Test, 2 agents | minMSD (K=12) | Improved |
Closest papers by meaning
- Scene Transformer: A unified architecture for predicting future trajectories of multiple agents (ICLR 2022 · similarity 0.91)
- Diverse and Admissible Trajectory Prediction through Multimodal Context Understanding (ECCV 2020 · similarity 0.88)
- Prediction by Anticipation: An Action-Conditional Prediction Method Based on Interaction Learning (ICCV 2021 · similarity 0.88)
- Motion Forecasting in Continuous Driving (NeurIPS 2024 · similarity 0.87)
- DeMo: Decoupling Motion Forecasting into Directional Intentions and Dynamic States (NeurIPS 2024 · similarity 0.86)
- Poly-Autoregressive Prediction for Modeling Interactions (CVPR 2025 · 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.
