Research map / Detection and segmentation / Open-set and few-shot detection
Human-Object Interaction Detection via Disentangled Transformer
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 |
|---|---|---|---|
| use disentangled human-object pair and interaction prediction with unified HOI features | DisTR · V-COCO test, Scenario #1 | mAP: 58.8 -> 66.2 | Helped |
| use disentangled human-object pair and interaction prediction with unified HOI features | DisTR · HICO-DET test, Default Full | mAP: 29.90 -> 31.75 | Helped |
| add encoder disentanglement for the different prediction tasks (ablation) | DisTR · V-COCO test, Scenario #1 | mAP: 65.5 -> 66.2 | Helped |
| add encoder disentanglement for the different prediction tasks (ablation) | DisTR · HICO-DET test, Default Full | mAP: 30.79 -> 31.75 | Helped |
| add attentional fusion between the instance and interaction decoders (ablation) | DisTR · V-COCO test, Scenario #1 | mAP: 64.4 -> 66.2 | Helped |
| add attentional fusion between the instance and interaction decoders (ablation) | DisTR · HICO-DET test, Default Full | mAP: 31.24 -> 31.75 | Helped |
| add decoder disentanglement and predict interactive human-object pairs (ablation) | DisTR · V-COCO test, Scenario #1 | mAP: 58.8 -> 66.2 | Helped |
| add decoder disentanglement and predict interactive human-object pairs (ablation) | DisTR · HICO-DET test, Default Full | mAP: 29.07 -> 31.75 | Helped |
| replace unified feature representation with learnable unified HOI queries (ablation) | DisTR · V-COCO test, Scenario #1 | mAP: 66.2 -> 64.9 | Hurt |
| replace unified feature representation with learnable unified HOI queries (ablation) | DisTR · HICO-DET test, Default Full | mAP: 31.75 -> 31.09 | Hurt |
| add warmup that initially freezes pretrained parameters (ablation) | DisTR · V-COCO test, Scenario #1 | mAP: 65.7 -> 66.2 | Helped |
| add warmup that initially freezes pretrained parameters (ablation) | DisTR · HICO-DET test, Default Full | mAP: 31.49 -> 31.75 | Helped |
Closest papers by meaning
- HOTR: End-to-End Human-Object Interaction Detection With Transformers (CVPR 2021 · similarity 0.92)
- End-to-End Human Object Interaction Detection With HOI Transformer (CVPR 2021 · similarity 0.92)
- MSTR: Multi-Scale Transformer for End-to-End Human-Object Interaction Detection (CVPR 2022 · similarity 0.92)
- Toward Open-Set Human Object Interaction Detection (AAAI 2024 · similarity 0.90)
- Dual-Prior Augmented Decoding Network for Long Tail Distribution in HOI Detection (AAAI 2024 · similarity 0.90)
- Exploring Pose-Aware Human-Object Interaction via Hybrid Learning (CVPR 2024 · 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.