Research map / Detection and segmentation / Weakly supervised detection and segmentation
Ambiguity-Resistant Semi-Supervised Learning for Dense Object 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 |
|---|---|---|---|
| Add JCE and TSA to FCOS | FCOS · COCO-Standard 10% | AP50:90: 30.7\% AP vs. 36.9\% AP | Helped |
| Add JCE and TSA to FCOS | FCOS · COCO-Standard | AP50:90: 22.82\pm0.26 vs. 22.43\pm0.36; 28.11\pm0.19 vs. 27.77\pm0.18; 33.14\pm0.12 vs. 32.50\pm0.08; 36.90\pm0.03 vs. 36.06\pm0. | Helped |
| Add ARSL semi-supervised training | FCOS · COCO-Full 100% | AP50:90: +4.70 | Helped |
| Add ARSL | FCOS · VOC2007 test | AP50: 71.36 to 80.40 | Helped |
| Add ARSL | FCOS · VOC2007 test | AP50:95: 45.52 to 56.40 | Helped |
| Add joint-confidence estimation (ablation) | Semi-supervised FCOS · COCO-Standard 10% | AP: 30.7 to 34.7 | Helped |
| Assign labels using joint confidence, without mining ambiguous candidates (ablation) | FCOS + JCE · COCO-Standard 10% | AP: 34.7 to 35.6 | Helped |
| Add potential-positive mining to TSA (ablation) | FCOS + JCE + TSA · COCO-Standard 10% | AP: 35.6 to 36.9 | Helped |
| Replace centerness estimation with IoU prediction (ablation) | Semi-supervised FCOS · COCO-Standard 10% | AP: 30.7 to 32.0; +1.3 | Helped |
| Train IoU prediction and classification with united supervision (ablation) | JCE · COCO-Standard 10% | AP: 32.0 to 34.2; +2.2 | Helped |
| Set unlabeled targets to the teacher's maximum responses (ablation) | JCE with united supervision · COCO-Standard 10% | AP: 34.2 to 34.7; +0.5 | Helped |
| Use JCE for pseudo-label selection | FCOS · COCO-Standard 10% | Top-5 IoU: 0.614 to 0.656 | Helped |
Closest papers by meaning
- Rethinking Pseudo Labels for Semi-supervised Object Detection (AAAI 2022 · similarity 0.92)
- Adapting Object Size Variance and Class Imbalance for Semi-supervised Object Detection (AAAI 2023 · similarity 0.91)
- Semi-DETR: Semi-Supervised Object Detection With Detection Transformers (CVPR 2023 · similarity 0.90)
- Dense Learning Based Semi-Supervised Object Detection (CVPR 2022 · similarity 0.90)
- Multi-clue Consistency Learning to Bridge Gaps Between General and Oriented Object in Semi-supervised Detection (AAAI 2025 · similarity 0.90)
- Sparse Semi-DETR: Sparse Learnable Queries for Semi-Supervised Object Detection (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.