Research map / Graph learning / Knowledge and temporal graphs
Improving Action Segmentation via Graph-Based Temporal Reasoning
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 GTRM to m-GRU | m-GRU · EGTEA | F1@10: 32.6 -> 41.6 | Helped |
| Add GTRM to m-GRU | m-GRU · EPIC-Kitchens | F1@10: 20.2 -> 31.9 | Helped |
| Add GTRM to m-GRU | m-GRU · EPIC-Kitchens | Edit: 30.5 -> 42.1 | Helped |
| Add GTRM to m-GRU | m-GRU · EGTEA | Acc: 67.1 -> 69.5 | Helped |
| Add GTRM to MSTCN | MSTCN · 50Salads | F1@10: 73.4 -> 75.4 | Helped |
| Add GTRM to Bi-LSTM | Bi-LSTM · 50Salads | F1@10: 62.2 -> 70.4 | Helped |
| Add GTRM to MSTCN with k = 2 | MSTCN · Breakfast | F1@10: 57.3 -> 57.5 | No clear effect |
| Add GTRM to MSTCN with k = 4 rather than k = 2 (ablation) | MSTCN · Breakfast | F1@10: 57.5 -> 57.3 | No clear effect |
| Use graph convolution for both classification and boundary regression rather than FCNs (ablation) | GTRM on m-GRU · EGTEA | F1@10 gain: 6.2 -> 10.0 | Helped |
| Use graph convolution for both classification and boundary regression rather than FCNs (ablation) | GTRM on m-GRU · EGTEA | Edit gain: 4.5 -> 7.5 | Helped |
| Include time vector ut (ablation) | GTRM on m-GRU · EGTEA | F1@10 gain: 4.6 -> 10.0 | Helped |
| Add GTRM to Bi-LSTM (ablation) | Bi-LSTM · 50Salads | F1@10: 62.2 -> 70.4 | Helped |
Closest papers by meaning
- Temporal Relational Modeling with Self-Supervision for Action Segmentation (AAAI 2021 · similarity 0.92)
- MS-TCN: Multi-Stage Temporal Convolutional Network for Action Segmentation (CVPR 2019 · similarity 0.89)
- G-TAD: Sub-Graph Localization for Temporal Action Detection (CVPR 2020 · similarity 0.89)
- FACT: Frame-Action Cross-Attention Temporal Modeling for Efficient Action Segmentation (CVPR 2024 · similarity 0.88)
- OnlineTAS: An Online Baseline for Temporal Action Segmentation (NeurIPS 2024 · similarity 0.87)
- Unsupervised Action Segmentation via Fast Learning of Semantically Consistent Actoms (AAAI 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.