Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Learning Predictive Models from Observation and Interaction
Karl Schmeckpeper, Annie Xie, Oleh Rybkin, Stephen Tian, Kostas Daniilidis, Sergey Levine, Chelsea Finn
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 action-free BDD100K driving video via POI | SAVP trained on Boston actions · nuScenes Singapore action-conditioned prediction (driving) | PSNR: 19.74 -> 20.88 | Improved |
| add action-free Singapore driving video via POI | SAVP trained on Boston actions · nuScenes Singapore action-conditioned prediction (driving) | PSNR: 19.74 -> 20.81 | Improved |
| add action-free Singapore video via POI | SAVP trained on Boston actions · nuScenes Singapore action-conditioned prediction (driving) | LPIPS: 0.1951 -> 0.1933 | Improved |
| add action-free BDD100K video via POI | SAVP trained on Boston actions · nuScenes Singapore action-conditioned prediction (driving) | LPIPS: 0.1951 -> 0.2106 | Hurt |
| use POI instead of shared inverse model | action-free Singapore video · nuScenes Singapore action-conditioned prediction (driving) | PSNR: 20.65 -> 20.81 | No clear effect |
| use POI instead of CLASP | action-free Singapore video · nuScenes Singapore action-conditioned prediction (driving) | PSNR: 20.57 -> 20.81 | No clear effect |
| add human tool-use video via POI | SAVP trained on random robot data · Robotic manipulation action-conditioned prediction (kinesthetic test set) | PSNR: 23.31 -> 23.79 | Improved |
| add human tool-use video via POI | SAVP trained on random robot data · Robotic manipulation action-conditioned prediction (kinesthetic test set) | SSIM: 0.803 -> 0.813 | Improved |
| add human tool-use video via POI | SAVP trained on random robot data · Robotic manipulation action-conditioned prediction (kinesthetic test set) | LPIPS: 0.0757 -> 0.0722 | Improved |
| use POI with domain-dependent prior instead of shared inverse model | human video plus random robot data · Robotic manipulation action-conditioned prediction (kinesthetic test set) | PSNR: 23.59 -> 23.79 | Improved |
| use CLASP instead of SAVP when adding human video | random robot data · Robotic manipulation action-conditioned prediction (kinesthetic test set) | PSNR: 23.31 -> 22.14 | Hurt |
| add human tool-use video via POI | SAVP in visual MPC · Robotic tool-use control (30 tasks, visual MPC) | success rate (%): 23.3% -> 40.0% | Improved |
Closest papers by meaning
- Predictive Inverse Dynamics Models are Scalable Learners for Robotic Manipulation (ICLR 2025 · similarity 0.89)
- Zero-Shot Robotic Manipulation with Pre-Trained Image-Editing Diffusion Models (ICLR 2024 · similarity 0.88)
- Robots Pre-train Robots: Manipulation-Centric Robotic Representation from Large-Scale Robot Datasets (ICLR 2025 · similarity 0.88)
- PlaySlot: Learning Inverse Latent Dynamics for Controllable Object-Centric Video Prediction and Planning (ICML 2025 · similarity 0.88)
- Behavioral Exploration: Learning to Explore via In-Context Adaptation (ICML 2025 · similarity 0.88)
- OSVI-WM: One-Shot Visual Imitation for Unseen Tasks using World-Model-Guided Trajectory Generation (NeurIPS 2025 · 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.
