Research map / Video / Video segmentation and tracking
Quantifying and Detecting Collective Motion by Manifold Learning
Qi Wang, Mulin Chen, Xuelong Li
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 the proposed manifold-based collectiveness instead of CT and MCC | Collectiveness measurement · Collective Motion Database (high/low, high/medium, medium/low) | precision, recall, F-measure | Improved |
| Replace MCC's manifold learning with the proposed neighbor-based manifold learning (ablation) | MCC · Collective Motion Database | precision, recall, F-measure | Mixed |
| Use the proposed multi-stage detection instead of CF, CT, CDC and MCC | Collective motion detection · CUHK Crowd Dataset | NMI, Purity, RI | Improved |
| Add global clustering refinement after local clustering (ablation) | Multi-stage clustering · CUHK Crowd Dataset (qualitative) | detection of global consistency | Improved |
| Set kNN parameter k to 20 rather than smaller or larger values (ablation) | Proposed method · Collective Motion Database classification | classification accuracy | Improved |
| Set manifold parameter alpha to 0.8 within 0.1-1 (ablation) | Proposed method · Collective Motion Database classification | classification accuracy | Improved |
| Exploit temporal dynamics via hidden-state model rather than frame-by-frame processing (CDC, MCC) | Collective motion detection · CUHK crowd videos, frames 15, 31, 48 | qualitative stability over time | Improved |
Closest papers by meaning
- Socially Constrained Structural Learning for Groups Detection in Crowd (TPAMI 2016 · similarity 0.87)
- DyCrowd: Towards Dynamic Crowd Reconstruction from a Large-scene Video (TPAMI 2025 · similarity 0.85)
- A Multiview-Based Parameter Free Framework for Group Detection (AAAI 2017 · similarity 0.85)
- Learning Extremely High Density Crowds as Active Matters (CVPR 2025 · similarity 0.84)
- Continuous Locomotive Crowd Behavior Generation (CVPR 2025 · similarity 0.84)
- Neural Persistence Dynamics (NeurIPS 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.
