Research map / Video / Video segmentation and tracking
Who's Better? Who's Best? Pairwise Deep Ranking for Skill Determination
Hazel Doughty, Dima Damen, Walterio Mayol-Cuevas
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 a similarity loss for pairs judged comparable in skill (ablation) | Siamese TSN with L_rank2 · Surgery | pairwise precision (%): 69.0 -> 70.2 | Improved |
| add a similarity loss for pairs judged comparable in skill (ablation) | Siamese TSN with L_rank2 · Dough-Rolling | pairwise precision (%): 78.5 -> 79.4 | Improved |
| add a similarity loss for pairs judged comparable in skill (ablation) | Siamese TSN with L_rank2 · Drawing | pairwise precision (%): 79.1 -> 83.2 | Improved |
| add a similarity loss for pairs judged comparable in skill (ablation) | Siamese TSN with L_rank2 · Chopstick-Using | pairwise precision (%): 68.8 -> 71.5 | Improved |
| add corresponding video splits to the ranking loss (ablation) | Siamese TSN with L_rank1 · Drawing | pairwise precision (%): 78.0 -> 79.1 | Improved |
| add corresponding video splits to the ranking loss (ablation) | Siamese TSN with L_rank1 · Chopstick-Using | pairwise precision (%): 68.8 -> 68.8 | No clear effect |
| use the proposed deep ranking method instead of the strongest baseline | Ours (L_rank3) · Surgery | pairwise precision (%): 66.1 -> 70.2 | Improved |
| use the proposed deep ranking method instead of the strongest baseline | Ours (L_rank3) · Dough-Rolling | pairwise precision (%): 78.1 -> 79.4 | Improved |
| use the proposed deep ranking method instead of the strongest baseline | Ours (L_rank3) · Drawing | pairwise precision (%): 72.0 -> 83.2 | Improved |
| use the proposed deep ranking method instead of the strongest baseline | Ours (L_rank3) · Chopstick-Using | pairwise precision (%): 76.6 -> 71.5 | Hurt |
| use the temporal stream rather than the spatial stream (ablation) | Siamese TSN with L_rank3 · Surgery | pairwise precision (%): 66.4 -> 72.5 | Improved |
| use the temporal stream rather than the spatial stream (ablation) | Siamese TSN with L_rank3 · Dough-Rolling | pairwise precision (%): 79.5 -> 79.5 | No clear effect |
Closest papers by meaning
- Learning Skill-Attributes for Transferable Assessment in Video (NeurIPS 2025 · similarity 0.83)
- SLVMBench: Skill Learning from Video Memory (NeurIPS 2026 · similarity 0.83)
- BASKET: A Large-Scale Video Dataset for Fine-Grained Skill Estimation (CVPR 2025 · similarity 0.82)
- Step Differences in Instructional Video (CVPR 2024 · similarity 0.82)
- LookWhen? Fast Video Recognition by Learning When, Where, and What to Compute (NeurIPS 2026 · similarity 0.82)
- ExAct: A Video-Language Benchmark for Expert Action Analysis (NeurIPS 2025 · similarity 0.82)
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.
