Research map / Architectures and efficiency / CNN design
Generating High-Quality Crowd Density Maps Using Contextual Pyramid CNNs
Vishwanath A. Sindagi, 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 |
|---|---|---|---|
| change DME architecture relative to Zhang et al. (deeper columns, different filters) (ablation) | Zhang et al. multi-column CNN · ShanghaiTech Part A | MAE: 110.2 -> 104.3 | Improved |
| change DME architecture relative to Zhang et al. (ablation) | Zhang et al. multi-column CNN · ShanghaiTech Part A | SSIM: 0.52 -> 0.54 | No clear effect |
| add global context estimator and F-CNN (ablation) | DME · ShanghaiTech Part A | MAE: 104.3 -> 89.9 | Improved |
| add global context estimator and F-CNN (ablation) | DME · ShanghaiTech Part A | MSE: 154.2 -> 127.9 | Improved |
| add global context estimator and F-CNN (ablation) | DME · ShanghaiTech Part A | SSIM: 0.54 -> 0.61 | Improved |
| add global context estimator and F-CNN (ablation) | DME · ShanghaiTech Part A | PSNR: 20.92 -> 20.97 | No clear effect |
| add local context estimator (ablation) | DME+GCE+FCNN · ShanghaiTech Part A | MAE: 89.9 -> 76.1 | Improved |
| add local context estimator (ablation) | DME+GCE+FCNN · ShanghaiTech Part A | MSE: 127.9 -> 110.2 | Improved |
| add local context estimator (ablation) | DME+GCE+FCNN · ShanghaiTech Part A | SSIM: 0.61 -> 0.65 | Improved |
| add local context estimator (ablation) | DME+GCE+FCNN · ShanghaiTech Part A | PSNR: 20.97 -> 21.4 | Improved |
| add adversarial loss to Euclidean loss (ablation) | DME+GCE+LCE+FCNN · ShanghaiTech Part A | MAE: 76.1 -> 73.6 | Improved |
| add adversarial loss to Euclidean loss (ablation) | DME+GCE+LCE+FCNN · ShanghaiTech Part A | SSIM: 0.65 -> 0.72 | Improved |
Closest papers by meaning
- Iterative Crowd Counting (ECCV 2018 · similarity 0.89)
- Adaptive Density Map Generation for Crowd Counting (ICCV 2019 · similarity 0.88)
- Context-Aware Crowd Counting (CVPR 2019 · similarity 0.87)
- Single-Image Crowd Counting via Multi-Column Convolutional Neural Network (CVPR 2016 · similarity 0.87)
- Kernel-Based Density Map Generation for Dense Object Counting (TPAMI 2020 · similarity 0.87)
- Composition Loss for Counting, Density Map Estimation and Localization in Dense Crowds (ECCV 2018 · 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.
