Research map / Video / Action recognition
Unsupervised Action Segmentation by Joint Representation Learning and Online Clustering
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 |
|---|---|---|---|
| Add a temporal order-preserving prior to standard OT (ablation) | Online clustering for activity segmentation · 50 Salads Eval | F1-Score: 27.8 -> 42.8 | Helped |
| Add a temporal order-preserving prior to standard OT (ablation) | Online clustering for activity segmentation · YouTube Instructions | F1-Score: 11.6 -> 30.0 | Helped |
| Add temporal coherence loss to TOT (ablation) | Temporal optimal transport segmentation · 50 Salads Eval | F1-Score: 42.8 -> 48.2 | Helped |
| Add temporal coherence loss to TOT (ablation) | Temporal optimal transport segmentation · 50 Salads Eval | MOF: 47.4 -> 44.5 | Hurt |
| Add temporal coherence loss to TOT (ablation) | Temporal optimal transport segmentation · YouTube Instructions | F1-Score: 30.0 -> 32.9 | Helped |
| Add CTE time-stamp prediction loss to TOT (ablation) | Temporal optimal transport segmentation · 50 Salads Eval | F1-Score: 42.8 -> 36.0 | Hurt |
| Add temporal coherence loss to TOT (ablation) | Temporal optimal transport segmentation · Breakfast | MOF | Hurt |
| Add temporal coherence loss to TOT | Temporal optimal transport segmentation · Desktop Assembly | F1-Score: 51.7 -> 53.4 | Helped |
| Use a joint online representation-learning and clustering framework rather than sequential clustering | Unsupervised activity segmentation · Breakfast | F1-Score: +4.6 | Helped |
| Add entropy regularization to the temporal OT objective (ablation) | TOT on 50 Salads Eval · 50 Salads Eval | MOF: 47.4% vs. 46.2% | Hurt |
Closest papers by meaning
- Temporally-Weighted Hierarchical Clustering for Unsupervised Action Segmentation (CVPR 2021 · similarity 0.92)
- Unsupervised Learning of Action Classes With Continuous Temporal Embedding (CVPR 2019 · similarity 0.89)
- Unsupervised Action Segmentation via Fast Learning of Semantically Consistent Actoms (AAAI 2024 · similarity 0.89)
- Temporally Consistent Unbalanced Optimal Transport for Unsupervised Action Segmentation (CVPR 2024 · similarity 0.88)
- Hierarchical Vector Quantization for Unsupervised Action Segmentation (AAAI 2025 · similarity 0.88)
- OnlineTAS: An Online Baseline for Temporal Action Segmentation (NeurIPS 2024 · similarity 0.87)
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.