Research map / Detection and segmentation / Open-set and few-shot detection
HOTR: End-to-End Human-Object Interaction Detection With Transformers
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 HOTR instead of the strongest listed prior method, IDN | V-COCO · V-COCO test Scenario 1 and Scenario 2 | AP_role: 53.3 -> 55.2; 60.3 -> 64.4 | Helped |
| use HOTR rather than IDN with a frozen COCO detector | HICO-DET · HICO-DET Default Full | mAP: 23.36 -> 23.46 | Helped |
| use HOTR rather than IDN with a frozen COCO detector | HICO-DET · HICO-DET Default Rare | mAP: 22.47 -> 16.21 | Hurt |
| use HOTR rather than IDN with a frozen COCO detector | HICO-DET · HICO-DET Default Non-Rare | mAP: 23.63 -> 25.62 | Helped |
| use HOTR rather than IDN with a fine-tuned detector | HICO-DET · HICO-DET Default Full | mAP: 24.58 -> 25.10 | Helped |
| use HOTR rather than IDN with a fine-tuned detector | HICO-DET · HICO-DET Default Non-Rare | mAP: 25.86 -> 27.42 | Helped |
| use HO Pointers instead of directly regressing human and object boxes (ablation) | HOTR · V-COCO test Scenario 1 | AP_role: 39.3 -> 55.2 | Helped |
| use HO Pointers instead of directly regressing human and object boxes (ablation) | HOTR · HICO-DET Default Full, COCO detector | mAP: 17.2 -> 23.5 | Helped |
| share the encoder across instance and interaction decoders (ablation) | HOTR · V-COCO test Scenario 1 | AP_role: 33.9 -> 55.2 | Helped |
| share the encoder across instance and interaction decoders (ablation) | HOTR · HICO-DET Default Full, COCO detector | mAP: 14.5 -> 23.5 | Helped |
| add interactiveness suppression for redundant human-object pairs (ablation) | HOTR · V-COCO test Scenario 1 | AP_role: 52.2 -> 55.2 | Helped |
| add interactiveness suppression for redundant human-object pairs (ablation) | HOTR · HICO-DET Default Full, COCO detector | mAP: 22.0 -> 23.5 | Helped |
Closest papers by meaning
- End-to-End Human Object Interaction Detection With HOI Transformer (CVPR 2021 · similarity 0.94)
- Human-Object Interaction Detection via Disentangled Transformer (CVPR 2022 · similarity 0.92)
- Exploring Structure-Aware Transformer Over Interaction Proposals for Human-Object Interaction Detection (CVPR 2022 · similarity 0.92)
- MSTR: Multi-Scale Transformer for End-to-End Human-Object Interaction Detection (CVPR 2022 · similarity 0.92)
- Efficient Two-Stage Detection of Human-Object Interactions With a Novel Unary-Pairwise Transformer (CVPR 2022 · similarity 0.91)
- 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.