Research map / Detection and segmentation / CNN features
Crowd Counting With Partial Annotations in an Image
Yanyu Xu, Ziming Zhong, Dongze Lian, Jing Li, Zhengxin Li, Xinxing Xu, Shenghua Gao
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 second density branch (Net1&2 multi-branch with sigma 15 and 20) (ablation) | CSRNet with 10% annotated regions · ShanghaiTech Part A | MAE: 83.87 -> 77.37 | Improved |
| add Unannotated Regions Characterization (VQ memory) (ablation) | Net1&2 · ShanghaiTech Part A | MAE: 77.37 -> 75.27 | Improved |
| add Cross-regressor Consistency Regularization (ablation) | Net1&2 · ShanghaiTech Part A | MAE: 77.37 -> 75.33 | Improved |
| add Feature Distribution Consistency (ablation) | Net1&2 · ShanghaiTech Part A | MAE: 77.37 -> 75.80 | Improved |
| combine URC, CCR and FDC (ablation) | Net1&2 · ShanghaiTech Part A | MAE: 77.37 -> 72.79 | Improved |
| combine URC, CCR and FDC (ablation) | Net1&2 · ShanghaiTech Part B | MAE: 13.51 -> 12.03 | Improved |
| use circle instead of rectangle annotated regions (ablation) | Ours · ShanghaiTech Part A | MAE: 72.79 -> 80.62 | Hurt |
| use triangle instead of rectangle annotated regions (ablation) | Ours · ShanghaiTech Part A | MAE: 72.79 -> 83.06 | Hurt |
| use circle or triangle instead of rectangle regions (ablation) | Ours · ShanghaiTech Part B | MAE: 12.03 -> 14.08 | Hurt |
| raise CCR loss weight from 0.1 to 1 (ablation) | Ours · ShanghaiTech Part A | MAE: 72.79 -> 136.08 | Hurt |
| raise mean/covariance loss weight from 0.01 to 0.1 (ablation) | Ours · ShanghaiTech Part A | MAE: 72.79 -> 73.43 | No clear effect |
| annotate 5% instead of 10% of each image | Ours · ShanghaiTech Part A | MAE: 72.79 -> 79.42 | Hurt |
Closest papers by meaning
- Active Crowd Counting with Limited Supervision (ECCV 2020 · similarity 0.90)
- Reducing Spatial Labeling Redundancy for Active Semi-Supervised Crowd Counting (TPAMI 2022 · similarity 0.89)
- Semi-Supervised Crowd Counting via Self-Training on Surrogate Tasks (ECCV 2020 · similarity 0.89)
- Modeling Noisy Annotations for Crowd Counting (NeurIPS 2020 · similarity 0.88)
- Counting without Scale: Scale-Consistent Error Correction for Crowd Counting (NeurIPS 2026 · similarity 0.88)
- CrowdCLIP: Unsupervised Crowd Counting via Vision-Language Model (CVPR 2023 · 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.
