Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Learning About Objects by Learning to Interact with Them
Martin Lohmann, Jordi Salvador, Aniruddha Kembhavi, Roozbeh Mottaghi
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 superpixel post-processing to self-supervised interaction masks (ablation) | Ours, segmentation only · NovelObjects | AP IoU=0.5: 12.74 -> 27.06 | Improved |
| Add superpixel post-processing to self-supervised interaction masks (ablation) | Ours, segmentation only · NovelSpaces | AP IoU=0.5: 13.24 -> 26.10 | Improved |
| Use prioritized memory-bank sampling (ablation) | Ours, segmentation only · NovelObjects | AP IoU=0.5: 21.04 -> 27.06 | Improved |
| Use prioritized memory-bank sampling (ablation) | Ours, segmentation only · NovelSpaces | AP IoU=0.5: 21.32 -> 26.10 | Improved |
| Limit the interaction range instead of using infinite arm length (ablation) | Ours, segmentation only · NovelObjects | AP IoU=0.5: 21.06 -> 27.06 | Improved |
| Limit the interaction range instead of using infinite arm length (ablation) | Ours, segmentation only · NovelSpaces | AP IoU=0.5: 20.10 -> 26.10 | Improved |
| Provide oracle ground-truth masks for selected interactions | Ours, segmentation only · NovelObjects | AP IoU=0.5: 27.06 -> 38.44 | Improved |
| Provide oracle ground-truth masks for selected interactions | Ours, segmentation only · NovelSpaces | AP IoU=0.5: 26.10 -> 36.49 | Improved |
| Provide oracle interaction locations for objects the model did not select | Ours, segmentation only · NovelSpaces | AP IoU=0.5: 26.10 -> 28.06 | Improved |
| Combine oracle masks and oracle interaction locations | Ours, segmentation only · NovelObjects | AP IoU=0.5: 27.06 -> 25.68 | No clear effect |
Closest papers by meaning
- Understanding 3D Object Interaction from a Single Image (ICCV 2023 · similarity 0.88)
- Reasoning About Physical Interactions with Object-Oriented Prediction and Planning (ICLR 2019 · similarity 0.87)
- The Curious Robot: Learning Visual Representations via Physical Interactions (ECCV 2016 · similarity 0.87)
- Physical Object Understanding with a Physically Controllable World Model (CVPR 2026 · similarity 0.86)
- Hierarchical Relational Inference (AAAI 2021 · similarity 0.86)
- Object Concepts Emerge from Motion (NeurIPS 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.
