Research map / Detection and segmentation / Open-set and few-shot detection
Affordance Transfer Learning for Human-Object Interaction Detection
Check 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 |
|---|---|---|---|
| Add affordance transfer using COCO object images | HOI detection on HICO-DET · HICO-DET Default Full | mAP: 23.63 -> 24.50 | No clear effect |
| Add ATL to DRG | DRG · HICO-DET Default Full | mAP: 24.53 -> 28.53 | Helped |
| Use COCO object images rather than HOI-COCO object images | ATL on HOI-COCO · HOI-COCO Rare | mAP: 8.01 -> 9.79 | Helped |
| Transfer affordances using HICO-DET object images | Novel-object HOI detection · HICO-DET novel object unseen category | mAP: 12.84 -> 11.35 | Hurt |
| Transfer affordances using COCO object images | Novel-object HOI detection · HICO-DET novel object unseen category | mAP: 12.84 -> 15.11 | Helped |
| Use two object images per batch instead of one (ablation) | ATL · HICO-DET Default Full | mAP: 24.07 -> 24.50 | Helped |
| Increase object images per batch from two to three (ablation) | ATL · HICO-DET Default Full | mAP: 24.50 -> 24.19 | Hurt |
| Use a COCO-pretrained detector rather than the baseline detector (ablation) | ATL · HICO-DET Default Full | mAP: 21.07 -> 20.08 | Hurt |
| Use COCO rather than HICO-DET object images (ablation) | ATL on HOI-COCO · HOI-COCO Full | mAP: 24.21 -> 24.74 | Helped |
| Extract verb representation from the human box rather than the union box (ablation) | ATL affordance recognition · COCO Val2017 | mAP: 52.01 -> 44.47 | Hurt |
| Use multiple object images per batch, increasing from two to three (ablation) | ATL affordance recognition · Object365 novel classes | F1: 12.86 -> 15.56 | Helped |
| Increase affordance templates from 10 to 100 (ablation) | ATL affordance recognition · HICO-DET | mAP: 59.14 -> 57.41 | Hurt |
Closest papers by meaning
- Weakly Supervised Multimodal Affordance Grounding for Egocentric Images (AAAI 2024 · similarity 0.89)
- Learning 2D Invariant Affordance Knowledge for 3D Affordance Grounding (AAAI 2025 · similarity 0.89)
- Exploring Pose-Aware Human-Object Interaction via Hybrid Learning (CVPR 2024 · similarity 0.88)
- Weakly-Supervised Affordance Grounding Guided by Part-Level Semantic Priors (ICLR 2025 · similarity 0.88)
- HORP: Human-Object Relation Priors Guided HOI Detection (CVPR 2025 · similarity 0.87)
- Open-World Human-Object Interaction Detection via Multi-modal Prompts (CVPR 2024 · 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.