Research map / Video / Action recognition
Boundary-Aware Cascade Networks for Temporal Action Segmentation
Zhenzhi Wang, Ziteng Gao, Limin Wang, Zhifeng Li, Gangshan Wu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 Stage Cascade to MS-TCN (ablation) | MS-TCN · 50Salads (mid) | Accuracy: 80.7 -> 83.4 | Improved |
| add Stage Cascade to MS-TCN (ablation) | MS-TCN · 50Salads (mid) | F1@50: 64.5 -> 48.9 | Hurt |
| add LBP post-processing to MS-TCN (ablation) | MS-TCN · 50Salads (mid) | F1@50: 64.5 -> 66.1 | Improved |
| add LBP post-processing to MS-TCN (ablation) | MS-TCN · 50Salads (mid) | Accuracy: 80.7 -> 81.5 | Improved |
| add LBP to Stage Cascade (ablation) | Stage Cascade · 50Salads (mid) | F1@50: 48.9 -> 74.0 | Improved |
| add LBP to Stage Cascade (ablation) | Stage Cascade · 50Salads (mid) | Accuracy: 83.4 -> 84.4 | Improved |
| pass video features to every MS-TCN stage (ablation) | MS-TCN · 50Salads (mid) | F1@50: 64.5 -> 45.8 | Hurt |
| replace attention with Stage Cascade plus LBP (ablation) | MS-TCN w/ attention&LBP · 50Salads (mid) | F1@50: 68.5 -> 74.0 | Improved |
| use BGM-based LBP instead of average smoothing (ablation) | BCN · 50Salads (mid) | F1@50: 69.1 -> 74.0 | Improved |
| use BGM-based LBP instead of Gaussian-like smoothing (ablation) | BCN · 50Salads (mid) | F1@50: 64.9 -> 74.0 | Improved |
| set threshold rho to 0.8 instead of 0.9 (ablation) | BCN · 50Salads (mid) | Accuracy: 82.9 -> 84.4 | Improved |
| use 3 cascade stages instead of 2 (ablation) | BCN · 50Salads (mid) | Accuracy: 83.6 -> 84.4 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- MS-TCN — MS-TCN: Multi-Stage Temporal Convolutional Network for Action Segmentation (CVPR 2019) · better on 6 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Fast and Unsupervised Action Boundary Detection for Action Segmentation (CVPR 2022 · similarity 0.89)
- Efficient and Effective Weakly-Supervised Action Segmentation via Action-Transition-Aware Boundary Alignment (CVPR 2024 · similarity 0.88)
- Temporal Convolutional Networks for Action Segmentation and Detection (CVPR 2017 · similarity 0.88)
- Efficient Temporal Action Segmentation via Boundary-aware Query Voting (NeurIPS 2024 · similarity 0.87)
- ActFusion: a Unified Diffusion Model for Action Segmentation and Anticipation (NeurIPS 2024 · similarity 0.86)
- FACT: Frame-Action Cross-Attention Temporal Modeling for Efficient Action Segmentation (CVPR 2024 · 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.
