Research map / Detection and segmentation / Weakly supervised detection and segmentation
Co-Student: Collaborating Strong and Weak Students for Sparsely Annotated Object Detection
Lianjun Wu, Jiangxiao Han, Zengqiang Zheng, Xinggang Wang
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 strong augmentation directly to the baseline FCOS (ablation) | FCOS · COCO-miss50p | AP: 33.5 -> 28.4 | Hurt |
| add denoising teacher model to dual-branch students (ablation) | Co-Student · COCO-miss50p | AP: 35.7 -> 36.1 | Improved |
| add strong augmentation to one student of dual branches (ablation) | Co-Student · COCO-miss50p | AP: 35.7 -> 36.4 | Improved |
| combine strong augmentation and denoising teacher (ablation) | Co-Student · COCO-miss50p | AP: 33.5 -> 37.0 | Improved |
| use dual training branches instead of single branch (ablation) | Teacher-student · COCO-miss50p | AP: 34.9 -> 35.7 | Improved |
| add strong augmentation in single training branch (ablation) | STB · COCO-miss50p | AP: 34.9 -> 34.4 | Hurt |
| add denoising teacher (ablation) | Co-Student · COCO-50p-val | false pseudo-label proportion | Improved |
| use Co-Student instead of Co-mining on deeper backbone | FCOS ResNeXt-32x8d-101-BiFPN · COCO-miss50p | AP: 40.8 -> 43.4 | Improved |
| Co-Student versus DenseTeacher (vanilla teacher-student) under same settings | Co-Student · COCO-miss50p | AP: 32.9 -> 37.0 | Improved |
| Co-Student versus Co-mining in the Extreme setting | Co-Student · COCO Extreme | AP: 23.1 -> 26.1 | Improved |
Later papers that compared with this method
1 later accepted paper reports Co-student as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- MonoSAOD: Monocular 3D Object Detection with Sparsely Annotated Label (CVPR 2026 · better on 1 of 1 shared result)
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.
- FCOS — FCOS: Fully Convolutional One-Stage Object Detection (ICCV 2019) · better on 4 of 4 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
- Calibrated Teacher for Sparsely Annotated Object Detection (AAAI 2023 · similarity 0.92)
- Co-mining: Self-Supervised Learning for Sparsely Annotated Object Detection (AAAI 2021 · similarity 0.92)
- SparseDet: Improving Sparsely Annotated Object Detection with Pseudo-positive Mining (ICCV 2023 · similarity 0.91)
- Label, Verify, Correct: A Simple Few Shot Object Detection Method (CVPR 2022 · similarity 0.91)
- Object Detection Benchmarks are Incomplete: The Role of Label Errors and Annotation Uncertainty (NeurIPS 2026 · similarity 0.89)
- As Pseudo-Label Free as Possible: Leveraging Adaptive Feature Generation for Sparsely Annotated Object Detection (AAAI 2025 · 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.
