Research map / Video / Action recognition
Temporally-Weighted Hierarchical Clustering for Unsupervised Action Segmentation
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 |
|---|---|---|---|
| Weight appearance distances by temporal distance in a hierarchical 1-nearest-neighbor graph | FINCH-style clustering · Breakfast | MoF: 51.9 -> 62.7 | Helped |
| Weight appearance distances by temporal distance in a hierarchical 1-nearest-neighbor graph | FINCH-style clustering · Inria Instructional Videos | MoF: 44.8 -> 56.7 | Helped |
| Weight appearance distances by temporal distance in a hierarchical 1-nearest-neighbor graph | FINCH-style clustering · 50Salads mid granularity | MoF: 33.7 -> 66.5 | Helped |
| Weight appearance distances by temporal distance in a hierarchical 1-nearest-neighbor graph | FINCH-style clustering · MPII Cooking 2 | MoF: 40.5 -> 42.0 | Helped |
| Weight appearance distances by temporal distance in a hierarchical 1-nearest-neighbor graph | FINCH-style clustering · Hollywood Extended | MoF: 56.8 -> 55.0 | Hurt |
| Set K to each video's ground-truth action count rather than the activity-average count | TW-FINCH · Breakfast | MoF: 62.7 -> 63.8 | Helped |
| Set K to each video's ground-truth action count rather than the activity-average count | TW-FINCH · Inria Instructional Videos | MoF: 56.7 -> 58.6 | Helped |
| Set K to each video's ground-truth action count rather than the activity-average count | TW-FINCH · 50Salads mid granularity | MoF: 66.5 -> 66.8 | Helped |
| Set K to each video's ground-truth action count rather than the activity-average count | TW-FINCH · MPII Cooking 2 | MoF: 42.0 -> 43.4 | Helped |
| Set K to each video's ground-truth action count rather than the activity-average count | TW-FINCH · Hollywood Extended | MoF: 55.0 -> 57.8 | Helped |
| Include all background frames instead of removing a ratio of background frames | TW-FINCH · Inria Instructional Videos | MoF: 56.7 -> 43.4 | Hurt |
Closest papers by meaning
- Unsupervised Action Segmentation by Joint Representation Learning and Online Clustering (CVPR 2022 · similarity 0.92)
- Unsupervised Action Segmentation via Fast Learning of Semantically Consistent Actoms (AAAI 2024 · similarity 0.92)
- Hierarchical Vector Quantization for Unsupervised Action Segmentation (AAAI 2025 · similarity 0.91)
- Iterative Contrast-Classify for Semi-supervised Temporal Action Segmentation (AAAI 2022 · similarity 0.91)
- FACT: Frame-Action Cross-Attention Temporal Modeling for Efficient Action Segmentation (CVPR 2024 · similarity 0.89)
- Temporally Consistent Unbalanced Optimal Transport for Unsupervised Action Segmentation (CVPR 2024 · similarity 0.88)
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.