Research map / Detection and segmentation / Open-set and few-shot detection
ViPLO: Vision Transformer Based Pose-Conditioned Self-Loop Graph for Human-Object Interaction Detection
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 |
|---|---|---|---|
| Replace the ResNet backbone with ViT and use the MOA feature extractor (ablation) | Two-stage HOI detection · HICO-DET Default Full | mAP: 29.26 -> 33.92 | Helped |
| Use overlapped-area MOA instead of ViT + ROI region extraction (ablation) | ViT-based HOI detection · HICO-DET Default Full | mAP: 32.84 -> 33.92 | Helped |
| Use overlapped-area masking instead of quantized patch masking (ablation) | ViT region feature extraction · HICO-DET Default Full | mAP: 33.20 -> 33.92 | Helped |
| Add human-pose information to edge encoding (ablation) | ViT + MOA graph · HICO-DET Default Full | mAP: 33.92 -> 34.51 | Helped |
| Add local human features extracted with ROIAlign to the pose-edge graph (ablation) | ViT + MOA graph · HICO-DET Default Full | mAP: 34.51 -> 34.95 | Helped |
| Extract local human features with MOA rather than ROIAlign (ablation) | Pose-edge graph with local human features · HICO-DET Default Full | mAP: 34.51 -> 34.44 | Hurt |
| Use a larger-patch ViT-B/32 backbone instead of ViT-B/16 (ablation) | ViPLO · HICO-DET Default Full | mAP: 36.97 -> 33.92 | Hurt |
| Use ViT-B/16 ViPLO rather than ViT-B/32 ViPLO | ViPLO · V-COCO Scenario 1 | AP_role: 60.9 -> 62.2 | Helped |
| Use ViT-B/16 ViPLO rather than ViT-B/32 ViPLO | ViPLO · V-COCO Scenario 2 | AP_role: 66.6 -> 68.0 | Helped |
| Replace CLIP-pretrained ViT with CLIP-pretrained ResNet50 (ablation) | ViPLO feature extractor · HICO-DET test Default | mAP: 32.23 -> 33.92 | Hurt |
Closest papers by meaning
- End-to-End Human Object Interaction Detection With HOI Transformer (CVPR 2021 · similarity 0.89)
- LINK: Learning Instance-level Knowledge from Vision-Language Models for Human-Object Interaction Detection (ICLR 2026 · similarity 0.89)
- Learning Human-Object Interaction Detection Using Interaction Points (CVPR 2020 · similarity 0.89)
- Exploring Structure-Aware Transformer Over Interaction Proposals for Human-Object Interaction Detection (CVPR 2022 · similarity 0.89)
- Exploring Pose-Aware Human-Object Interaction via Hybrid Learning (CVPR 2024 · similarity 0.88)
- End-to-End HOI Reconstruction Transformer with Graph-based Encoding (CVPR 2025 · similarity 0.86)
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.