Research map / Adversarial robustness and security / Attacks on text and graphs
Crowd Counting via Adversarial Cross-Scale Consistency Pursuit
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 adversarial and perceptual losses to Euclidean-loss training (ablation) | ACSCP · ShanghaiTech Part A | MAE: 95.8 -> 83.2 | Helped |
| Add adversarial and perceptual losses to Euclidean-loss training (ablation) | ACSCP · ShanghaiTech Part B | MAE: 24.1 -> 18.4 | Helped |
| Add adversarial and perceptual losses to Euclidean-loss training (ablation) | ACSCP · WorldExpo’10 | MAE: 9.95 -> 8.48 | Helped |
| Add cross-scale consistency loss to the integrated adversarial and perceptual objective (ablation) | ACSCP · ShanghaiTech Part A | MAE: 83.2 -> 75.7 | Helped |
| Add cross-scale consistency loss to the integrated adversarial and perceptual objective (ablation) | ACSCP · ShanghaiTech Part B | MAE: 18.4 -> 17.2 | Helped |
| Add cross-scale consistency loss to the integrated adversarial and perceptual objective (ablation) | ACSCP · WorldExpo’10 | MAE: 8.48 -> 7.5 | Helped |
| Use ACSCP instead of CP-CNN | crowd counting · ShanghaiTech Part A | MAE: 73.6 -> 75.7 | Hurt |
| Use ACSCP instead of CP-CNN | crowd counting · UCF CC 50 | MSE: 320.9 -> 404.6 | Hurt |
| Use ACSCP instead of MCNN | crowd counting · UCSD | MSE: 1.35 -> 1.35 | No clear effect |
| Vary the cross-scale consistency loss weight (ablation) | ACSCP on ShanghaiTech Part B · ShanghaiTech Part B | MAE: 18.4, 17.8, 17.3, 17.2, 17.7, 22.5 | Mixed |
Closest papers by meaning
- Counting without Scale: Scale-Consistent Error Correction for Crowd Counting (NeurIPS 2026 · similarity 0.90)
- Generative Adversarial Perturbations with Cross-paradigm Transferability on Localized Crowd Counting (CVPR 2026 · similarity 0.90)
- Residual Regression With Semantic Prior for Crowd Counting (CVPR 2019 · similarity 0.90)
- Crowd Counting and Density Estimation by Trellis Encoder-Decoder Networks (CVPR 2019 · similarity 0.90)
- Context-Aware Crowd Counting (CVPR 2019 · similarity 0.89)
- Free Lunch Enhancements for Multi-modal Crowd Counting (CVPR 2025 · 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.