Research map / Label-efficient and robust learning / Semi-supervised learning
Learning to Count in the Crowd from Limited Labeled Data
Vishwanath A. Sindagi, Rajeev Yasarla, Deepak Sam Babu, R. Venkatesh Babu, Vishal M. Patel
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 GP pseudo-GT supervision on unlabeled data | ResNet-50 counter · SH-A (5% labeled) | MAE: 118 -> 102 | Improved |
| add GP pseudo-GT supervision on unlabeled data | ResNet-50 counter · UCF-QNRF (5% labeled) | MAE: 186 -> 160 | Improved |
| use GP pseudo-GT instead of ranking loss on unlabeled data | ResNet-50 counter · SH-A (5% labeled) | MAE: 115 -> 102 | Improved |
| add ranking loss on unlabeled data | ResNet-50 counter · UCSD (5% labeled) | MAE: 2.2 -> 2.2 | No clear effect |
| increase labeled data from 5% to 75% with GP (ablation) | ResNet-50 counter · SH-A | MAE: 102 -> 88 | Improved |
| add GP unlabeled data at 75% labeled (ablation) | ResNet-50 counter · SH-A (75% labeled) | MAE: 93 -> 88 | Improved |
| add GP unlabeled data at 25% labeled (ablation) | ResNet-50 counter · UCF-QNRF (25% labeled) | MAE: 178 -> 147 | Improved |
| add GP with ResNet-101 encoder (ablation) | ResNet-101 counter · SH-A (5% labeled) | MAE: 131 -> 110 | Improved |
| add GP with VGG16 encoder (ablation) | VGG16 counter · UCF-QNRF (5% labeled) | MAE: 188 -> 175 | Improved |
| add GP with CSRNet encoder (ablation) | CSRNet · SH-A (5% labeled) | MAE: 120 -> 111 | Improved |
| add GP with Res101-SFCN encoder (ablation) | Res101-SFCN · SH-A (5% labeled) | MAE: 128 -> 109 | Improved |
| raise unsupervised loss weight from 0 to 0.6 (ablation) | GP method · SH-A val (5% labeled) | MAE: 102 -> 85 | Improved |
Closest papers by meaning
- Active Crowd Counting with Limited Supervision (ECCV 2020 · similarity 0.90)
- Completely Self-Supervised Crowd Counting via Distribution Matching (ECCV 2022 · similarity 0.89)
- Semi-Supervised Crowd Counting via Self-Training on Surrogate Tasks (ECCV 2020 · similarity 0.88)
- Semi-Supervised Counting via Pixel-by-Pixel Density Distribution Modeling (TPAMI 2025 · similarity 0.88)
- CrowdDiff: Multi-hypothesis Crowd Density Estimation using Diffusion Models (CVPR 2024 · similarity 0.86)
- Taste More, Taste Better: Diverse Data and Strong Model Boost Semi-Supervised Crowd Counting (CVPR 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.
