Research map / Detection and segmentation / CNN features
DecideNet: Counting Varying Density Crowds Through Attention Guided Detection and Density Estimation
Read the paperCheck 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 |
|---|---|---|---|
| Combine detection and regression density estimates using learned attention | DecideNet · Mall | MAE: 2.01 -> 1.52 | Helped |
| Combine detection and regression density estimates using learned attention | DecideNet · WorldExpo'10 (Ave) | MAE: 9.40 -> 9.23 | Helped |
| Add an R3 stream | DecideNet · ShanghaiTech PartB | MAE: 21.53 -> 20.75 | Helped |
| Add an R3 stream | DecideNet · ShanghaiTech PartB | MSE: 31.98 -> 29.42 | Helped |
| Add detection and attention to RegNet-only counting (ablation) | DecideNet · Mall | MAE: 3.37 -> 1.52 | Helped |
| Add detection and attention to RegNet-only counting (ablation) | DecideNet · ShanghaiTech PartB | MAE: 42.85 -> 21.53 | Helped |
| Add regression and attention to DetNet-only counting (ablation) | DecideNet · Mall | MAE: 4.50 -> 1.52 | Helped |
| Add regression and attention to DetNet-only counting (ablation) | DecideNet · ShanghaiTech PartB | MAE: 44.90 -> 21.53 | Helped |
| Replace learned attention with late averaging fusion (ablation) | DecideNet · Mall | MAE: 1.52 -> 3.93 | Hurt |
| Replace learned attention with late averaging fusion (ablation) | DecideNet · ShanghaiTech PartB | MAE: 21.53 -> 38.63 | Hurt |
| Add QualityNet attention rather than use late fusion (ablation) | DecideNet · Mall | MAE: 3.93 -> 1.83 | Helped |
| Add QualityNet attention rather than use late fusion (ablation) | DecideNet · ShanghaiTech PartB | MAE: 38.63 -> 24.93 | Helped |
Closest papers by meaning
- Attention Scaling for Crowd Counting (CVPR 2020 · similarity 0.90)
- Shallow Feature Based Dense Attention Network for Crowd Counting (AAAI 2020 · similarity 0.90)
- Crowd Counting and Density Estimation by Trellis Encoder-Decoder Networks (CVPR 2019 · similarity 0.89)
- Wide-Area Crowd Counting via Ground-Plane Density Maps and Multi-View Fusion CNNs (CVPR 2019 · similarity 0.89)
- Single Domain Generalization for Crowd Counting (CVPR 2024 · similarity 0.87)
- CrowdDiff: Multi-hypothesis Crowd Density Estimation using Diffusion Models (CVPR 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.