Research map / Video / Action recognition
Context-Aware and Scale-Insensitive Temporal Repetition Counting
Huaidong Zhang, Xuemiao Xu, Guoqiang Han, Shengfeng He
Check 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 |
|---|---|---|---|
| Use the proposed context-aware, scale-insensitive framework | Temporal repetition counting · QUVA Repetition | MAE: 0.232 ± 0.344 -> 0.148 ± 0.290 | Improved |
| Use the proposed context-aware, scale-insensitive framework | Temporal repetition counting · YTsegments | MAE: 0.065 ± 0.092 -> 0.053 ± 0.115 | Improved |
| Use the proposed context-aware, scale-insensitive framework | Temporal repetition counting · UCFRep | MAE: 0.286 ± 0.574 -> 0.147 ± 0.243 | Improved |
| Use five refinement stages instead of three (ablation) | Coarse-to-fine refinement · UCFRep validation | MAE: 0.157 ± 0.284 -> 0.147 ± 0.243 | Improved |
| Use six rather than five refinement stages (ablation) | Coarse-to-fine refinement · UCFRep validation | MAE: 0.147 ± 0.243 -> 0.151 ± 0.254 | Hurt |
| Predict separate time-scales for consecutive cycles instead of sharing one (ablation) | Context-aware estimation network · UCFRep validation | MAE: 0.177 ± 0.280 -> 0.157 ± 0.243 | Improved |
| Add multi-anchors to separate cycle time-scales (ablation) | Context-aware estimation network · UCFRep validation | MAE: 0.157 ± 0.243 -> 0.147 ± 0.243 | Improved |
| Add multi-anchors while sharing a single time-scale (ablation) | Context-aware estimation network · UCFRep validation | MAE: 0.177 ± 0.280 -> 0.171 ± 0.249 | Improved |
| Add multi-anchors while sharing a single time-scale (ablation) | Context-aware estimation network · UCFRep validation | OBOA: 0.70 -> 0.71 | Improved |
Closest papers by meaning
- Counting Out Time: Class Agnostic Video Repetition Counting in the Wild (CVPR 2020 · similarity 0.93)
- Efficient Action Counting with Dynamic Queries (IJCV 2026 · similarity 0.91)
- TransRAC: Encoding Multi-Scale Temporal Correlation With Transformers for Repetitive Action Counting (CVPR 2022 · similarity 0.91)
- CountLLM: Towards Generalizable Repetitive Action Counting via Large Language Model (CVPR 2025 · similarity 0.90)
- Learning to Predict Activity Progress by Self-Supervised Video Alignment (CVPR 2024 · similarity 0.84)
- OnlineTAS: An Online Baseline for Temporal Action Segmentation (NeurIPS 2024 · similarity 0.83)
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.
