Research map / Architectures and efficiency / CNN design
Attention Scaling for Crowd Counting
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 |
|---|---|---|---|
| add attention scaling with 2 density masks (ablation) | ASNet · ShanghaiTech Part A | MAE: 68.31 -> 60.16 | Helped |
| add 2-level Adaptive Pyramid Loss instead of MSE loss (ablation) | ASNet · ShanghaiTech Part A | MAE: 60.16 -> 57.78 | Helped |
| use 3-level rather than 2-level Adaptive Pyramid Loss (ablation) | ASNet · ShanghaiTech Part A | MAE: 57.78 -> 58.99 | Hurt |
| remove scaling factors while retaining 2 attention masks (ablation) | ASNet · ShanghaiTech Part A | MAE: 60.16 -> 62.70 | Hurt |
| use 4 instead of 2 attention masks with scaling (ablation) | ASNet · ShanghaiTech Part A | MAE: 60.16 -> 61.44 | Hurt |
| use 8 instead of 2 attention masks with scaling (ablation) | ASNet · ShanghaiTech Part A | MAE: 60.16 -> 61.78 | Hurt |
Closest papers by meaning
- Shallow Feature Based Dense Attention Network for Crowd Counting (AAAI 2020 · similarity 0.92)
- Boosting Crowd Counting via Multifaceted Attention (CVPR 2022 · similarity 0.91)
- Top-Down Feedback for Crowd Counting Convolutional Neural Network (AAAI 2018 · similarity 0.90)
- DecideNet: Counting Varying Density Crowds Through Attention Guided Detection and Density Estimation (CVPR 2018 · similarity 0.90)
- Single-Image Crowd Counting via Multi-Column Convolutional Neural Network (CVPR 2016 · similarity 0.90)
- Gramformer: Learning Crowd Counting via Graph-Modulated Transformer (AAAI 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.