Research map / Image generation / Image restoration
Single-Image Crowd Counting via Multi-Column Convolutional Neural Network
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 |
|---|---|---|---|
| Use three columns with different receptive-field sizes rather than a single column (ablation) | MCNN · Shanghaitech Part A test | MAE: 141.2, 160.5, 153.7 -> 110.2 | Helped |
| Use three columns with different receptive-field sizes rather than a single column (ablation) | MCNN · Shanghaitech Part A test | MSE: 206.8, 239.0, 230.2 -> 173.2 | Helped |
| Add per-column pretraining before joint fine-tuning (ablation) | MCNN · Shanghaitech Part A test | MAE: 122.8 -> 110.2 | Helped |
| Add per-column pretraining before joint fine-tuning (ablation) | MCNN · Shanghaitech Part A test | MSE: 185.9 -> 173.2 | Helped |
| Estimate a density map instead of predicting total image count directly (ablation) | MCNN · Shanghaitech Part A test | MAE: 245.0 -> 110.2 | Helped |
| Estimate a density map instead of predicting total image count directly (ablation) | MCNN · Shanghaitech Part B test | MAE: 70.9 -> 26.4 | Helped |
| Fine-tune only the last two layers rather than the whole network (ablation) | MCNN transferred from Shanghaitech Part A · UCF CC 50 | MAE: 378.3 -> 295.1 | Helped |
| Transfer the source-trained MCNN without target-domain training | MCNN trained on Shanghaitech Part A · UCF CC 50 | MAE: 377.7 -> 397.7 | Hurt |
| Fine-tune the whole network rather than only the last two layers (ablation) | MCNN transferred from Shanghaitech Part A · UCF CC 50 | MAE: 295.1 -> 378.3 | Hurt |
Closest papers by meaning
- Switching Convolutional Neural Network for Crowd Counting (CVPR 2017 · similarity 0.93)
- Top-Down Feedback for Crowd Counting Convolutional Neural Network (AAAI 2018 · similarity 0.92)
- Wide-Area Crowd Counting via Ground-Plane Density Maps and Multi-View Fusion CNNs (CVPR 2019 · similarity 0.91)
- Context-Aware Crowd Counting (CVPR 2019 · similarity 0.91)
- Revisiting Perspective Information for Efficient Crowd Counting (CVPR 2019 · similarity 0.90)
- Gramformer: Learning Crowd Counting via Graph-Modulated Transformer (AAAI 2024 · similarity 0.87)
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.