Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Learning to Act Properly: Predicting and Explaining Affordances From Images
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 |
|---|---|---|---|
| Use spatially adjacent object edges instead of unary object features | Affordance relationship prediction · Sit, ground-truth segmentation | mAcc: 0.730 -> 0.745 | Helped |
| Use spatially adjacent object edges instead of unary object features | Affordance relationship prediction · Run, ground-truth segmentation | mAcc: 0.714 -> 0.730 | Helped |
| Use spatially adjacent object edges instead of unary object features | Affordance relationship prediction · Grasp, ground-truth segmentation | mAcc: 0.594 -> 0.604 | Hurt |
| Use spatial GGNN context rather than unary object features | Explanation generation · Run, ground-truth segmentation | CIDEr: 0.350 -> 0.409 | Helped |
| Use spatial GGNN context rather than unary object features | Consequence generation · Grasp, ground-truth segmentation | CIDEr: 0.772 -> 1.504 | Helped |
| Add object class information (ablation) | Spatial GGNN · Sit, ground-truth segmentation | mAcc: 0.622 -> 0.745 | Helped |
| Add object ResNet features (ablation) | Spatial GGNN · Sit, ground-truth segmentation | mAcc: 0.720 -> 0.745 | Helped |
| Add global image features (ablation) | Spatial GGNN · Sit, ground-truth segmentation | mAcc: 0.720 -> 0.745 | Helped |
| Use T=3 propagation steps instead of T=1 (ablation) | Spatial GGNN · Sit, ground-truth segmentation | mAcc: 0.723 -> 0.745 | Helped |
| Use T=3 propagation steps instead of T=4 (ablation) | Spatial GGNN · Sit, ground-truth segmentation | mAcc: 0.738 -> 0.745 | Helped |
| Share GGNN parameters across relationship, explanation, and consequence tasks (ablation) | Single-action multi-task model · Sit, Run, and Grasp | mixed: null | Mixed |
| Share GGNN parameters across tasks and actions (ablation) | Multi-action multi-task model · Sit, Run, and Grasp | mixed: null | Mixed |
Closest papers by meaning
- Weakly Supervised Multimodal Affordance Grounding for Egocentric Images (AAAI 2024 · similarity 0.90)
- Probing and Bridging Geometry-Interaction Cues for Affordance Reasoning in Vision Foundation Models (CVPR 2026 · similarity 0.90)
- Affordance-R1: Reinforcement Learning for Generalizable Affordance Reasoning in Multimodal Large Language Models (AAAI 2026 · similarity 0.89)
- GREAT: Geometry-Intention Collaborative Inference for Open-Vocabulary 3D Object Affordance Grounding (CVPR 2025 · similarity 0.88)
- AffordBot: 3D Fine-grained Embodied Reasoning via Multimodal Large Language Models (NeurIPS 2025 · similarity 0.88)
- IAAO: Interactive Affordance Learning for Articulated Objects in 3D Environments (CVPR 2025 · similarity 0.88)
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.