Research map / Detection and segmentation / Open-set and few-shot detection
Mining Instance-Centric Vision-Language Contexts for Human-Object Interaction Detection
Soo Won Seo, KyungChae Lee, Hyungchan Cho, Taein Son, Nam Ik Cho, Jun Won Choi
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 ICR module to the detector-only baseline (ablation) | InCoM-Net · HICO-DET Full | mAP: 36.17 -> 37.42 | Improved |
| add ProCA after ICR (ablation) | InCoM-Net · HICO-DET Full | mAP: 37.42 -> 38.42 | Improved |
| add masked feature training (ablation) | InCoM-Net · HICO-DET Full | mAP: 38.42 -> 39.53 | Improved |
| use global, intra-instance and inter-instance contexts instead of raw VLM features (ablation) | ICR · HICO-DET Full | mAP: 38.30 -> 39.53 | Improved |
| add global context only to raw VLM features (ablation) | ICR · HICO-DET Rare | mAP: 37.31 -> 36.76 | Hurt |
| add intra-instance context (ablation) | ICR · HICO-DET Rare | mAP: 36.76 -> 38.78 | Improved |
| increase ProCA layers from 1 to 4 (ablation) | ProCA · HICO-DET Full | mAP: 37.42 -> 38.48 | Improved |
| add masked feature training to reduce VLM over-reliance (ablation) | InCoM-Net · HICO-DET Full, D-only inference | mAP: 19.11 -> 36.14 | Improved |
| add masked feature training | HOICLIP · HICO-DET Full | mAP: 34.00 -> 35.41 | Improved |
| use ICR+ProCA instead of RoI Align (ablation) | InCoM-Net · HICO-DET Full | mAP: 36.37 -> 38.42 | Improved |
| use ICR+ProCA instead of image crop (ablation) | InCoM-Net · HICO-DET Full | mAP: 37.34 -> 38.42 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- HORP — HORP: Human-Object Relation Priors Guided HOI Detection (CVPR 2025) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- LINK: Learning Instance-level Knowledge from Vision-Language Models for Human-Object Interaction Detection (ICLR 2026 · similarity 0.93)
- Learning Human-Object Interaction Detection Using Interaction Points (CVPR 2020 · similarity 0.91)
- ContextHOI: Spatial Context Learning for Human-Object Interaction Detection (AAAI 2025 · similarity 0.91)
- Learning Human-Object Interaction as Groups (NeurIPS 2025 · similarity 0.90)
- Exploring the Potential of Large Foundation Models for Open-Vocabulary HOI Detection (CVPR 2024 · similarity 0.90)
- Bilateral Adaptation for Human-Object Interaction Detection with Occlusion-Robustness (CVPR 2024 · similarity 0.90)
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.
