Research map / Detection and segmentation / Open-set and few-shot detection
Exploring Structure-Aware Transformer Over Interaction Proposals 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 hard mining of high-scoring negative pairs (ablation) | Base interaction proposal network · V-COCO | AP_role#1: 52.49 -> 58.45 | Helped |
| Add hard mining of high-scoring negative pairs (ablation) | Base interaction proposal network · HICO-DET Default Full | mAP: 21.74 -> 24.16 | Helped |
| Add vanilla Transformer relational reasoning (ablation) | Base+HM · V-COCO | AP_role#1: 58.45 -> 63.50 | Helped |
| Add vanilla Transformer relational reasoning (ablation) | Base+HM · HICO-DET Default Full | mAP: 24.16 -> 28.62 | Helped |
| Add inter-interaction structure-aware self-attention (ablation) | Base+HM+TR · V-COCO | AP_role#1: 63.50 -> 64.99 | Helped |
| Add inter-interaction structure-aware self-attention (ablation) | Base+HM+TR · HICO-DET Default Full | mAP: 28.62 -> 29.65 | Helped |
| Add intra-interaction structure-aware cross-attention (ablation) | Base+HM+TR · V-COCO | AP_role#1: 63.50 -> 65.04 | Helped |
| Add intra-interaction structure-aware cross-attention (ablation) | Base+HM+TR · HICO-DET Default Full | mAP: 28.62 -> 29.74 | Helped |
| Use both structure-aware self-attention and cross-attention (ablation) | Base+HM+TR · V-COCO | AP_role#1: 63.50 -> 66.04 | Helped |
| Use both structure-aware self-attention and cross-attention (ablation) | Base+HM+TR · HICO-DET Default Full | mAP: 28.62 -> 30.56 | Helped |
| Use interaction proposals as non-parametric queries and encode inter-/intra-interaction structure | STIP · V-COCO | AP_role#1: 58.8 -> 66.0 | Helped |
| Use interaction proposals as non-parametric queries and encode inter-/intra-interaction structure | STIP · HICO-DET Default Full, detector pre-trained on MS-COCO | mAP: 24.40 -> 28.81 | Helped |
Closest papers by meaning
- End-to-End Human Object Interaction Detection With HOI Transformer (CVPR 2021 · similarity 0.94)
- What To Look at and Where: Semantic and Spatial Refined Transformer for Detecting Human-Object Interactions (CVPR 2022 · similarity 0.92)
- HOTR: End-to-End Human-Object Interaction Detection With Transformers (CVPR 2021 · similarity 0.92)
- Human-Object Interaction Detection via Disentangled Transformer (CVPR 2022 · similarity 0.92)
- Exploring Pose-Aware Human-Object Interaction via Hybrid Learning (CVPR 2024 · similarity 0.90)
- An Image-like Diffusion Method for Human-Object Interaction Detection (CVPR 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.