Research map / Detection and segmentation / Open-set and few-shot detection
Semantics-Guided Contrastive Network for Zero-Shot Object Detection
Caixia Yan, Xiaojun Chang, Minnan Luo, Huan Liu, Xiaoqin Zhang, Qinghua Zheng
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 region-region contrastive learning (RRCL) (ablation) | ContrastZSD · PASCAL VOC ZSD | mAP: 61.5 -> 65.7 | Improved |
| add region-region contrastive learning (RRCL) (ablation) | ContrastZSD · PASCAL VOC GZSD | harmonic mean mAP: 50.6 -> 53.6 | Improved |
| add unseen-class contrastive learning (RCCL_u) (ablation) | ContrastZSD · PASCAL VOC ZSD | mAP: 61.2 -> 65.7 | Improved |
| add unseen-class contrastive learning (RCCL_u) (ablation) | ContrastZSD · PASCAL VOC GZSD unseen | mAP: 30.6 -> 48.3 | Improved |
| add unseen-class contrastive learning (RCCL_u) (ablation) | ContrastZSD · PASCAL VOC GZSD | harmonic mean mAP: 40.8 -> 53.6 | Improved |
| mapping-contrastive strategy vs best prior ZSD (PL) | ContrastZSD · PASCAL VOC ZSD | mAP: 62.1 -> 65.7 | Improved |
| contrastive model vs HRE on GZSD unseen | ContrastZSD · PASCAL VOC GZSD | mAP: 37.0 -> 46.5 | Improved |
| contrastive model vs ConSE-ZSD on seen-class detection | ContrastZSD · PASCAL VOC seen | mAP: 77.0 -> 76.7 | No clear effect |
| ContrastZSD vs BLC, 48/17 split ZSD | ContrastZSD · MS COCO 48/17 | Recall@100: 48.8 -> 52.4 | Improved |
| ContrastZSD vs BLC, 65/15 split ZSD | ContrastZSD · MS COCO 65/15 | mAP: 14.7 -> 18.6 | Improved |
| ContrastZSD vs BLC, GZSD 48/17 | ContrastZSD · MS COCO GZSD 48/17 | harmonic mean Recall@100: 51.4 -> 58.3 | Improved |
| ContrastZSD vs BLC, GZSD 65/15 | ContrastZSD · MS COCO GZSD 65/15 | harmonic mean mAP: 19.2 -> 23.4 | Improved |
Closest papers by meaning
- Zero-Shot Object Detection (ECCV 2018 · similarity 0.93)
- Meta-ZSDETR: Zero-shot DETR with Meta-learning (ICCV 2023 · similarity 0.91)
- Transductive Learning for Zero-Shot Object Detection (ICCV 2019 · similarity 0.91)
- Improved Visual-Semantic Alignment for Zero-Shot Object Detection (AAAI 2020 · similarity 0.90)
- Transferable Contrastive Network for Generalized Zero-Shot Learning (ICCV 2019 · similarity 0.89)
- ZeroMamba: Exploring Visual State Space Model for Zero-Shot Learning (AAAI 2025 · similarity 0.88)
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.
