Research map / Graph learning / GNN architectures
Hybrid Graph Neural Networks for Crowd Counting
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 |
|---|---|---|---|
| Add hybrid graph reasoning for multi-scale and cross-domain relations (ablation) | truncated VGG crowd-counting baseline · ShanghaiTech Part A | MAE: 68.2↦ 60.2 | Helped |
| Add localization cooperation to single-task graph modeling (ablation) | Single-task GNN · ShanghaiTech Part A | MAE: 62.5↦ 60.2 | Helped |
| Add the learnable cross-domain adapter (ablation) | Multi-task GNN · ShanghaiTech Part A | MAE: 62.4 vs. 60.2 | Helped |
| Increase nodes per domain from 2 to 3 (ablation) | HyGnn · ShanghaiTech Part A | MAE: 62.1↦ 60.2 | Helped |
| Increase nodes per domain from 3 to 5 (ablation) | HyGnn · ShanghaiTech Part A | MAE: 60.1↦ 60.2 | Mixed |
| Increase message-passing iterations from 1 to 3 (ablation) | HyGnn · ShanghaiTech Part A | MAE: 65.4↦60.2 | Helped |
| Increase message-passing iterations from 3 to 5 (ablation) | HyGnn · ShanghaiTech Part A | MAE: 60.2↦60.1 | Mixed |
| Use graph-based multi-scale aggregation instead of PSP aggregation (ablation) | Baseline + PSP · ShanghaiTech Part A | MAE: 65.3 -> 62.5 | Helped |
| Use graph-based multi-scale aggregation instead of bidirectional fusion (ablation) | Baseline + Bidirectional Fusion · ShanghaiTech Part A | MAE: 65.1 -> 62.5 | Helped |
Closest papers by meaning
- Gramformer: Learning Crowd Counting via Graph-Modulated Transformer (AAAI 2024 · similarity 0.88)
- Attention Scaling for Crowd Counting (CVPR 2020 · similarity 0.86)
- Switching Convolutional Neural Network for Crowd Counting (CVPR 2017 · similarity 0.86)
- Leveraging Heterogeneous Auxiliary Tasks to Assist Crowd Counting (CVPR 2019 · similarity 0.86)
- Free Lunch Enhancements for Multi-modal Crowd Counting (CVPR 2025 · similarity 0.86)
- Gaussian Density Splatting Network (NeurIPS 2026 · similarity 0.84)
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.