Research map / Detection and segmentation / CNN features
Fast and Unsupervised Action Boundary Detection for 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 |
|---|---|---|---|
| detect boundaries from smoothed-frame similarity and refine segments | ABD · Breakfast | MoF: 62.7 -> 64.0 | Helped |
| detect boundaries from smoothed-frame similarity and refine segments | ABD · YouTube Instructional Videos | MoF: 56.7 -> 67.2 | Helped |
| detect boundaries from smoothed-frame similarity and refine segments | ABD · Hollywood Extended | MoF: 55.0 -> 60.7 | Helped |
| detect boundaries from smoothed-frame similarity and refine segments | ABD · 50Salads mid | MoF: 66.5 -> 71.8 | Helped |
| add feature smoothing (ablation) | ABD · Breakfast | MoF: 58.3 -> 64.0 | Helped |
| use Gaussian rather than average feature smoothing (ablation) | ABD · Breakfast | MoF: 63.5 -> 64.0 | No clear effect |
| add non-maximum suppression (ablation) | ABD · Breakfast | MoF: 57.3 -> 64.0 | Helped |
| add segment clustering refinement (ablation) | ABD · Breakfast | MoF: 57.8 -> 64.0 | Helped |
| include temporal information in refinement (ablation) | ABD · Breakfast | MoF: 64.0 -> 64.2 | No clear effect |
| include temporal information in refinement (ablation) | ABD · YouTube Instructional Videos and Hollywood Extended | MoF: 67.2 -> 66.4; 60.7 -> 60.5 | Hurt |
| include temporal information in refinement (ablation) | ABD · 50Salads mid | MoF: 71.8 -> 72.2 | No clear effect |
| use I3D rather than Fisher Vector features (ablation) | ABD · Breakfast | MoF: 64.0 -> 65.4 | Helped |
Closest papers by meaning
- Unsupervised Action Segmentation via Fast Learning of Semantically Consistent Actoms (AAAI 2024 · similarity 0.91)
- Efficient and Effective Weakly-Supervised Action Segmentation via Action-Transition-Aware Boundary Alignment (CVPR 2024 · similarity 0.90)
- Temporally-Weighted Hierarchical Clustering for Unsupervised Action Segmentation (CVPR 2021 · similarity 0.89)
- Efficient Temporal Action Segmentation via Boundary-aware Query Voting (NeurIPS 2024 · similarity 0.87)
- Hierarchical Vector Quantization for Unsupervised Action Segmentation (AAAI 2025 · similarity 0.86)
- Understanding Multi-Task Activities from Single-Task Videos (CVPR 2025 · similarity 0.86)
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.