Research map / Video / Video segmentation and tracking
DR.VIC: Decomposition and Reasoning for Video Individual 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 |
|---|---|---|---|
| Use DRNet decomposition and differentiable optimal transport instead of the strongest tracking baseline | Video individual counting · CroHD test | MAE: 253.2 -> 141.1 | Helped |
| Use DRNet instead of the strongest tracking baseline | Video individual counting · CroHD test | MSE: 300.8 -> 192.3 | Helped |
| Use DRNet instead of the strongest baseline | Video individual counting · CroHD test | MRAE (%): 32.7 -> 27.4 | Helped |
| Use DRNet instead of LOI | Video individual counting · SenseCrowd test | MAE: 24.7 -> 12.3 | Helped |
| Use DRNet instead of LOI | Video individual counting · SenseCrowd test | MSE: 33.1 -> 24.7 | Helped |
| Use DRNet instead of LOI | Video individual counting · SenseCrowd test | MRAE (%): 37.4 -> 12.7 | Helped |
| Use differentiable optimal transport rather than tracking-based association (ablation) | DRNet · CroHD test | MAE: 526.4 -> 141.1 | Helped |
| Use differentiable optimal transport rather than the Hungarian algorithm (ablation) | DRNet · CroHD test | MAE: 313.4 -> 141.1 | Helped |
| Add predicted head proposals during training alongside ground-truth points (ablation) | DRNet · CroHD test | MAE: 251.2 -> 141.1 | Helped |
| Add predicted head proposals during training alongside ground-truth points (ablation) | DRNet · CroHD test | MRAE (%): 54.1 -> 27.4 | Helped |
| Add hard-negative pair loss (ablation) | DRNet · CroHD test | MAE: 189.4 -> 141.1 | Helped |
| Add hard-negative pair loss (ablation) | DRNet · CroHD test | MRAE (%): 38.2 -> 27.4 | Helped |
Closest papers by meaning
- Weakly Supervised Video Individual Counting (CVPR 2024 · similarity 0.86)
- Point in, Box Out: Beyond Counting Persons in Crowds (CVPR 2019 · similarity 0.85)
- Learning to Count via Unbalanced Optimal Transport (AAAI 2021 · similarity 0.85)
- Point-to-Region Loss for Semi-Supervised Point-Based Crowd Counting (CVPR 2025 · similarity 0.84)
- Single Domain Generalization for Crowd Counting (CVPR 2024 · similarity 0.83)
- Referring Expression Counting (CVPR 2024 · similarity 0.83)
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.