Research map / Detection and segmentation / Weakly supervised detection and segmentation
Multi-Task Self-Supervised Object Detection via Recycling of Bounding Box Annotations
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 |
|---|---|---|---|
| Add the three bounding-box-recycled auxiliary tasks | Faster R-CNN with ResNet-101 · VOC07 test, VOC07+12 trainval | mAP: 81.7 -> 83.7 | Helped |
| Add the three bounding-box-recycled auxiliary tasks | Faster R-CNN with ResNet-101 · VOC12 test, VOC07+12 trainval | mAP: 75.3 -> 77.5 | Helped |
| Add the three bounding-box-recycled auxiliary tasks | Faster R-CNN with ResNet-101 · COCO17 test-dev | mAP@[.5:.95]: 32.8 -> 34.7 | Helped |
| Add multi-object labeling (ablation) | Faster R-CNN with ResNet-101 · VOC12 test, VOC07+12 trainval | mAP: 75.3 -> 77.4 | Helped |
| Add closeness labeling (ablation) | Faster R-CNN with ResNet-101 · VOC12 test, VOC07+12 trainval | mAP: 75.3 -> 76.0 | Helped |
| Add foreground labeling (ablation) | Faster R-CNN with ResNet-101 · VOC12 test, VOC07+12 trainval | mAP: 75.3 -> 75.1 | No clear effect |
| Add multi-object and closeness labeling together (ablation) | Faster R-CNN with ResNet-101 · VOC12 test, VOC07+12 trainval | mAP: 75.3 -> 77.3 | Helped |
| Add foreground labeling to multi-object and closeness labeling (ablation) | Faster R-CNN with ResNet-101 · VOC12 test, VOC07+12 trainval | mAP: 77.3 -> 77.5 | Helped |
| Add auxiliary-task feature learning without refinement (ablation) | Faster R-CNN with ResNet-101 · VOC07 test, VOC07 trainval | mAP: 77.0 -> 78.0 (+1.0) | Helped |
| Add detection refinement without auxiliary-task feature learning (ablation) | Faster R-CNN with ResNet-101 · VOC07 test, VOC07 trainval | mAP: 77.0 -> 78.3 (+1.3) | Helped |
Closest papers by meaning
- Open-Vocabulary Object Detection Using Captions (CVPR 2021 · similarity 0.88)
- We Don't Need No Bounding-Boxes: Training Object Class Detectors Using Only Human Verification (CVPR 2016 · similarity 0.87)
- Instance Localization for Self-Supervised Detection Pretraining (CVPR 2021 · similarity 0.87)
- Large Scale Semi-Supervised Object Detection Using Visual and Semantic Knowledge Transfer (CVPR 2016 · similarity 0.87)
- Exploit Bounding Box Annotations for Multi-Label Object Recognition (CVPR 2016 · similarity 0.87)
- Self-Supervised Learning of Intertwined Content and Positional Features for Object Detection (ICML 2025 · similarity 0.86)
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.