Research map / 3D vision / Depth and stereo
Calibration-Free Multi-View Crowd Counting
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 confidence maps to weight map (ablation) | CF-MVCC · CVCS | MAE: 16.46 -> 13.90 | Improved |
| add image-feature confidence only (ablation) | CF-MVCC · CVCS | MAE: 16.46 -> 16.13 | No clear effect |
| add distance-based confidence only (ablation) | CF-MVCC · CVCS | MAE: 16.46 -> 16.12 | No clear effect |
| combine image and distance confidence (ablation) | CF-MVCC · CVCS | MAE: 16.12 -> 13.90 | Improved |
| use LCC single-view network instead of CSRNet (ablation) | CF-MVCC-C · CVCS | MAE: 13.90 -> 12.79 | Improved |
| use ground-truth homography instead of predicted (ablation) | CF-MVCC-C · CVCS | MAE: 13.90 -> 11.69 | Improved |
| replace homography VPM with 4D correlation | CF-MVCC · CVCS | MAE: 16.46 -> 17.76 | Hurt |
| replace pairwise matching with total count loss only | CVCS · CVCS | MAE: 16.46 -> 18.89 | Hurt |
| increase test views from 5 to 9 (trained on 5) (ablation) | CF-MVCC-C · CVCS | MAE: 13.90 -> 22.23 | Hurt |
| add unsupervised domain adaptation (ablation) | CF-MVCC-C · CityStreet | MAE: 48.58 -> 14.89 | Improved |
| add one labeled view fine-tuning of SVC (ablation) | CF-MVCC-C · CityStreet | MAE: 14.89 -> 8.13 | Improved |
| add one labeled view fine-tuning of SVC (ablation) | CF-MVCC-C · PETS2009 | MAE: 12.76 -> 4.85 | Improved |
Closest papers by meaning
- Cross-View Cross-Scene Multi-View Crowd Counting (CVPR 2021 · similarity 0.91)
- 3D Crowd Counting via Multi-View Fusion with 3D Gaussian Kernels (AAAI 2020 · similarity 0.91)
- Wide-Area Crowd Counting via Ground-Plane Density Maps and Multi-View Fusion CNNs (CVPR 2019 · similarity 0.91)
- Wide-Area Crowd Counting: Multi-view Fusion Networks for Counting in Large Scenes (IJCV 2022 · similarity 0.90)
- 3D Crowd Counting via Geometric Attention-Guided Multi-view Fusion (IJCV 2022 · similarity 0.90)
- CountFormer: Multi-View Crowd Counting Transformer (ECCV 2024 · 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.
