Research map / Detection and segmentation / Open-set and few-shot detection
Detecting and Recognizing Human-Object Interactions
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 |
|---|---|---|---|
| Add person-conditioned, action-specific target localization (ablation) | InteractNet · V-COCO val | AP_role: 31.9 -> 37.5 | Helped |
| Add person-conditioned, action-specific target localization (ablation) | InteractNet · V-COCO test | AP_role: 32.6 -> 40.0 | Helped |
| Add the interaction branch to score actions using both human and object appearance (ablation) | InteractNet · V-COCO val | AP_role: 36.8 -> 37.5 | Helped |
| Use ResNet-50-FPN instead of vanilla ResNet-50 (ablation) | InteractNet · V-COCO val | AP_role: 35.9 -> 37.5 | Helped |
| Replace pairwise logit summation with concatenated features and a two-layer MLP (ablation) | InteractNet · V-COCO val | AP_role: 37.5 -> 37.5 | No clear effect |
| Replace target-location regression with a one-component mixture density network (ablation) | InteractNet · V-COCO val | AP_role: 37.5 -> 37.7 | No clear effect |
| Replace target-location regression with a two-component mixture density network (ablation) | InteractNet · V-COCO val | AP_role: 37.5 -> 37.3 | No clear effect |
| Use InteractNet rather than the reimplemented Gupta & Malik model C | V-COCO test (all actions) · V-COCO test (all actions) | AP_role: 31.8 -> 40.0 | Helped |
| Use InteractNet rather than the published approach | HICO-DET · HICO-DET test (full) | mAP: 7.81 -> 9.94 | Helped |
Closest papers by meaning
- Exploring Self- and Cross-Triplet Correlations for Human-Object Interaction Detection (AAAI 2024 · similarity 0.89)
- Exploring Pose-Aware Human-Object Interaction via Hybrid Learning (CVPR 2024 · similarity 0.88)
- Learning Human-Object Interaction as Groups (NeurIPS 2025 · similarity 0.88)
- Visual Diversity and Region-aware Prompt Learning for Zero-shot HOI Detection (NeurIPS 2025 · similarity 0.87)
- InteractAnything: Zero-shot Human Object Interaction Synthesis via LLM Feedback and Object Affordance Parsing (CVPR 2025 · similarity 0.87)
- HORP: Human-Object Relation Priors Guided HOI Detection (CVPR 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.