Research map / 3D vision / Depth and stereo
Wide-Area Crowd Counting: Multi-view Fusion Networks for Counting in Large Scenes
Qi Zhang, Antoni B. Chan
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 projection normalization to late fusion (ablation) | late fusion · PETS2009 scene | MAE: 4.22 -> 3.92 | Improved |
| add projection normalization to late fusion (ablation) | late fusion · CityStreet scene | MAE: 9.87 -> 8.12 | Improved |
| add projection normalization to late fusion (ablation) | late fusion · DukeMTMC scene | MAE: 1.43 -> 1.27 | Improved |
| use learnable scale selection instead of fixed discrete (ablation) | MVMS · PETS2009 scene | MAE: 3.82 -> 3.49 | Improved |
| use learnable scale selection instead of fixed discrete (ablation) | MVMS · CityStreet scene | MAE: 7.80 -> 8.01 | Hurt |
| add scale selection to naive early fusion (MVMS) | early fusion · PETS2009 scene | MAE: 5.42 -> 3.49 | Improved |
| add rotation selection module (MVMSR vs MVMS) (ablation) | MVMS · CityStreet scene | MAE: 7.36 -> 6.98 | Improved |
| add rotation selection module (MVMSR vs MVMS) (ablation) | MVMS · DukeMTMC scene | MAE: 0.95 -> 0.89 | Improved |
| add rotation selection module (MVMSR vs MVMS) (ablation) | MVMS · PETS2009 scene | MAE: 3.49 -> 3.62 | Hurt |
| use soft mask instead of discrete mask for fixed scale selection (ablation) | MVMS · CityStreet scene | MAE: 7.80 -> 8.55 | Hurt |
| use larger CSR-net backbone instead of FCN-7 (ablation) | MVMS · CityStreet scene | MAE: 8.01 -> 7.36 | Improved |
| use larger backbone (CSR-net) instead of FCN-7 (ablation) | MVMS · PETS2009 scene | MAE: 3.49 -> 3.99 | Hurt |
Closest papers by meaning
- Wide-Area Crowd Counting via Ground-Plane Density Maps and Multi-View Fusion CNNs (CVPR 2019 · similarity 0.99)
- 3D Crowd Counting via Multi-View Fusion with 3D Gaussian Kernels (AAAI 2020 · similarity 0.95)
- 3D Crowd Counting via Geometric Attention-Guided Multi-view Fusion (IJCV 2022 · similarity 0.95)
- Multi-Level Bottom-Top and Top-Bottom Feature Fusion for Crowd Counting (ICCV 2019 · similarity 0.93)
- Cross-View Cross-Scene Multi-View Crowd Counting (CVPR 2021 · similarity 0.91)
- Switching Convolutional Neural Network for Crowd Counting (CVPR 2017 · similarity 0.91)
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.
