Research map / Video / Action recognition
Anchor-Constrained Viterbi for Set-Supervised 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 |
|---|---|---|---|
| Use anchor-constrained Viterbi instead of set-constrained Viterbi | Set-supervised action segmentation · Breakfast | Mof: 30.2 -> 33.4 | Helped |
| Use anchor-constrained Viterbi instead of set-constrained Viterbi | Set-supervised action segmentation · Cooking 2 | midpoint: 14.5 -> 15.5 | Helped |
| Use anchor-constrained Viterbi instead of set-constrained Viterbi | Set-supervised action segmentation · Hollywood Extended | IoD: 17.7 -> 20.9 | Helped |
| Use anchor-constrained Viterbi instead of set-constrained Viterbi | Action alignment with known test action sets · Breakfast | Mof: 40.8 -> 45.1 | Helped |
| Use anchor-constrained Viterbi instead of set-constrained Viterbi | Action alignment with known test action sets · Cooking 2 | midpoint: 15.1 -> 16.2 | Helped |
| Use anchor-constrained Viterbi instead of set-constrained Viterbi | Action alignment with known test action sets · Hollywood Extended | IoD: 35.5 -> 40.8 | Helped |
| Add updated HMM transition probabilities, mean lengths, and class priors (ablation) | ACV action segmentation · Breakfast | Mof: 31.9 -> 33.4 | Helped |
| Add diversity regularization (ablation) | ACV action segmentation · Breakfast | Mof: 32.1 -> 33.4 | Helped |
| Add diversity regularization instead of N-pair loss (ablation) | ACV action segmentation · Hollywood Extended | IoD: 20.2 -> 20.9 | Helped |
| Set anchor-segment length parameter α to 0.6 rather than 0.4, 0.8, or 1.0 (ablation) | ACV action segmentation · Breakfast | Mof: 32.5 -> 33.4; 31.7 -> 33.4; 30.5 -> 33.4 | Helped |
| Set anchor-segment length parameter α to 0.6 rather than 1.0 (ablation) | Salient anchor estimation during training · Breakfast | IoD: 18.1 -> 19.4; 25.1 -> 28.7; 29.4 -> 32.2; 33.7 -> 36.6; 35.2 -> 39.0; 35.9 -> 39.2 | Helped |
Closest papers by meaning
- Set-Constrained Viterbi for Set-Supervised Action Segmentation (CVPR 2020 · similarity 0.95)
- Action Sets: Weakly Supervised Action Segmentation Without Ordering Constraints (CVPR 2018 · similarity 0.89)
- NeuralNetwork-Viterbi: A Framework for Weakly Supervised Video Learning (CVPR 2018 · similarity 0.88)
- Efficient and Effective Weakly-Supervised Action Segmentation via Action-Transition-Aware Boundary Alignment (CVPR 2024 · similarity 0.88)
- FACT: Frame-Action Cross-Attention Temporal Modeling for Efficient Action Segmentation (CVPR 2024 · similarity 0.86)
- 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.