Research map / Architectures and efficiency / CNN design
Crowd Counting and Density Estimation by Trellis Encoder-Decoder Networks
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 |
|---|---|---|---|
| Replace the single-path decoder with a multi-path trellis decoder (ablation) | TEDnet · ShanghaiTech Part A | MAE: 73.1 -> 71.2; 2.6% | Helped |
| Replace the single-path decoder with a multi-path trellis decoder (ablation) | TEDnet · ShanghaiTech Part A | PSNR: 22.51 -> 24.24; 7.1% | Helped |
| Add distributed supervision (ablation) | TEDnet with multi-path trellis decoder · ShanghaiTech Part A | MAE: 71.2 -> 69.2; 2.8% | Helped |
| Add distributed supervision (ablation) | TEDnet with multi-path trellis decoder · ShanghaiTech Part A | PSNR: 24.24 -> 24.71; 1.9% | Helped |
| Replace normal MSE with SAL (ablation) | TEDnet with distributed supervision · ShanghaiTech Part A | MAE: 69.2 -> 67.8 | Helped |
| Add SCL to SAL (ablation) | TEDnet with distributed supervision · ShanghaiTech Part A | MAE: 69.2 -> 64.2; 7.2% | Helped |
| Add SCL to SAL (ablation) | TEDnet with distributed supervision · ShanghaiTech Part A | PSNR: 24.71 -> 25.88; 4.5% | Helped |
| Use TEDnet instead of SANet | Crowd counting · ShanghaiTech Part A | MAE: 4.2% | Helped |
| Use TEDnet instead of SANet | Crowd counting · ShanghaiTech Part B | MSE: 5.9% | Helped |
| Use TEDnet instead of the second-best approach | Crowd counting · UCF-CC-50 | MAE: 3.5% | Helped |
| Use TEDnet instead of the second-best approach | Crowd counting · UCF-QNRF | MAE: 14.4% | Helped |
| Use TEDnet instead of the second-best approach | Crowd counting · UCF-QNRF | MSE: 1.6% | Helped |
Closest papers by meaning
- Wide-Area Crowd Counting via Ground-Plane Density Maps and Multi-View Fusion CNNs (CVPR 2019 · similarity 0.90)
- Almost Unsupervised Learning for Dense Crowd Counting (AAAI 2019 · similarity 0.90)
- CrowdCLIP: Unsupervised Crowd Counting via Vision-Language Model (CVPR 2023 · similarity 0.90)
- Switching Convolutional Neural Network for Crowd Counting (CVPR 2017 · similarity 0.90)
- Crowd Counting via Adversarial Cross-Scale Consistency Pursuit (CVPR 2018 · similarity 0.90)
- Gramformer: Learning Crowd Counting via Graph-Modulated Transformer (AAAI 2024 · similarity 0.89)
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.