Research map / Video / Video segmentation and tracking
Temporal Rate Reduction Clustering for Human Motion Segmentation
Xianghan Meng, Zhengyu Tong, Zhiyuan Huang, Chun-Guang 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 |
|---|---|---|---|
| add temporal regularizer L_r (ablation) | TR2C · Weiz (HoG, ablation) | ACC (%): 37.30 -> 94.07 | Improved |
| add temporal regularizer L_r (ablation) | TR2C · Keck (HoG, ablation) | ACC (%): 47.29 -> 86.78 | Improved |
| add temporal regularizer L_r (ablation) | TR2C · UT (HoG, ablation) | ACC (%): 45.79 -> 94.05 | Improved |
| add total coding rate term L_rho (ablation) | TR2C · Weiz (HoG, ablation) | ACC (%): 53.14 -> 94.07 | Improved |
| add total coding rate term L_rho (ablation) | TR2C · MAD (HoG, ablation) | ACC (%): 50.54 -> 83.99 | Improved |
| add cluster coding rate term L_rho^c (ablation) | TR2C · Weiz (HoG, ablation) | ACC (%): 64.68 -> 94.07 | Improved |
| add cluster coding rate term L_rho^c (ablation) | TR2C · YouTube (VGG, ablation) | ACC (%): 48.16 -> 96.40 | Improved |
| remove temporal regularizer on YouTube (ablation) | TR2C · YouTube (VGG, ablation) | ACC (%): 94.82 -> 96.40 | No clear effect |
| replace HoG features with CLIP features | TR2C · Keck | ACC (%): 83.50 -> 90.10 | Improved |
| replace HoG features with CLIP features | TR2C · Weiz | ACC (%): 94.12 -> 96.21 | Improved |
| replace HoG features with CLIP features | TSC · Weiz | ACC (%): 61.11 -> 89.61 | Improved |
| use learned TR2C representations instead of raw HoG for spectral clustering | spectral clustering · UT (HoG) | clustering accuracy: 29% improvement | Improved |
Closest papers by meaning
- Learning Transferable Subspace for Human Motion Segmentation (AAAI 2018 · similarity 0.90)
- Graph Constrained Data Representation Learning for Human Motion Segmentation (ICCV 2021 · similarity 0.90)
- Probabilistic Temporal Subspace Clustering (CVPR 2017 · similarity 0.88)
- Learning Segmentation from Point Trajectories (NeurIPS 2024 · similarity 0.86)
- Learned Trajectory Embedding for Subspace Clustering (CVPR 2024 · similarity 0.85)
- Hierarchical Vector Quantization for Unsupervised Action Segmentation (AAAI 2025 · similarity 0.85)
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.
