Research map / Detection and segmentation / Open-set and few-shot detection
Where2Explore: Few-shot Affordance Learning for Unseen Novel Categories of Articulated Objects
Chuanruo Ning, Ruihai Wu, Haoran Lu, Kaichun Mo, Hao Dong
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 |
|---|---|---|---|
| use similarity-guided interactions on novel categories | Where2Explore · Novel categories, pushing and pulling | F-score: 35.4 / 38.5 / 41.6; 12.1 / 12.5 / 24.2 | Improved |
| use similarity-guided interactions on novel categories | Where2Explore · Novel categories, pushing and pulling | Sample successful rate: 31.3 / 37.3 / 39.5; 11.5 / 13.4 / 14.9 | Improved |
| add similarity-guided few-shot exploration (ablation) | No-explore affordance model · Novel categories | F-score: 20.6; 3.9 | Improved |
| replace similarity-guided exploration with random interactions (ablation) | Where2Explore · Novel categories, pushing and pulling | F-score: 24.9 / 28.8 / 29.0; 4.0 / 6.1 / 9.2 | Hurt |
| replace similarity-guided exploration with random interactions (ablation) | Where2Explore · Novel categories, pushing and pulling | Sample successful rate: 15.0 / 18.4 / 21.6; 3.4 / 5.2 / 5.1 | Hurt |
| use Where2Act-style exploration instead of the similarity module (ablation) | Where2Explore · Novel categories, pushing and pulling | F-score: 24.8 / 25.2 / 30.2; 6.8 / 6.7 / 8.4 | Hurt |
| use Where2Act-style exploration instead of the similarity module (ablation) | Where2Explore · Novel categories, pushing and pulling | Sample successful rate: 15.0 / 19.2 / 25.5; 5.4 / 6.2 / 5.3 | Hurt |
| remove few-shot exploration (ablation) | Where2Explore · Novel categories | F-score: 20.6; 3.9 | Hurt |
| train on all categories with abundant data (ablation) | Where2Explore · Novel categories | F-score: 47.9; 27.1 | Improved |
| use similarity-guided few-shot exploration rather than abundant full-category training (ablation) | Where2Explore · Novel categories | Sample successful rate: 39.5 vs 40.3; 14.9 vs 13.9 | Mixed |
Closest papers by meaning
- Adaptive Articulated Object Manipulation On The Fly with Foundation Model Reasoning and Part Grounding (ICCV 2025 · similarity 0.92)
- AdaAfford: Learning to Adapt Manipulation Affordance for 3D Articulated Objects via Few-Shot Interactions (ECCV 2022 · similarity 0.92)
- AffordDP: Generalizable Diffusion Policy with Transferable Affordance (CVPR 2025 · similarity 0.91)
- IAAO: Interactive Affordance Learning for Articulated Objects in 3D Environments (CVPR 2025 · similarity 0.90)
- VidBot: Learning Generalizable 3D Actions from In-the-Wild 2D Human Videos for Zero-Shot Robotic Manipulation (CVPR 2025 · similarity 0.90)
- BiAssemble: Learning Collaborative Affordance for Bimanual Geometric Assembly (ICML 2025 · similarity 0.89)
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.
