Research map / Label-efficient and robust learning / Recognition and long-tail
Uncertainty-Aware Score Distribution Learning for Action Quality Assessment
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 |
|---|---|---|---|
| Replace single-score regression with Gaussian score-distribution learning | I3D AQA model · AQA-7 | Average Spearman correlation: 0.7472 -> 0.8102 | Helped |
| Use MUSDL with judge-score paths and ground-truth difficulty degree | I3D AQA model · MTL-AQA | Spearman correlation: 0.8905 -> 0.9273 | Helped |
| Add difficulty degree to score-distribution learning (ablation) | USDL · MTL-AQA | Spearman correlation: 0.9066 -> 0.9231 | Helped |
| Add multiple judge-score paths to the single-path method (ablation) | USDL_DD · MTL-AQA | Spearman correlation: 0.9231 -> 0.9273 | Helped |
| Predict difficulty degree instead of using its ground-truth test-time value (ablation) | MUSDL · MTL-AQA | Spearman correlation: 0.9273 -> 0.9158 | Hurt |
| Replace Gaussian score labels with χ² distributions (ablation) | USDL · AQA-7 | Average Spearman correlation: 0.8102 -> 0.8015 | Hurt |
| Replace Gaussian score labels with triangle distributions (ablation) | USDL · AQA-7 | Average Spearman correlation: 0.8102 -> 0.7969 | Hurt |
| Use Gaussian rather than alternative soft score distributions (ablation) | USDL · AQA-7 action classes | Spearman correlation | Mixed |
| Use 10 segments with s1 frame indices rather than the 6-segment scheme (ablation) | USDL · AQA-7 Diving | Spearman correlation: 0.7642 -> 0.8099 | Helped |
| Use 10-seg-s1 rather than 10-seg-s2 frame indices (ablation) | USDL · AQA-7 Diving | Spearman correlation: 0.7928 -> 0.8099 | Helped |
Closest papers by meaning
- LucidAction: A Hierarchical and Multi-model Dataset for Comprehensive Action Quality Assessment (NeurIPS 2024 · similarity 0.87)
- FineParser: A Fine-grained Spatio-temporal Action Parser for Human-centric Action Quality Assessment (CVPR 2024 · similarity 0.85)
- GAIA: Rethinking Action Quality Assessment for AI-Generated Videos (NeurIPS 2024 · similarity 0.84)
- PTM-VQA: Efficient Video Quality Assessment Leveraging Diverse PreTrained Models from the Wild (CVPR 2024 · similarity 0.83)
- Teaching Large Language Models to Regress Accurate Image Quality Scores Using Score Distribution (CVPR 2025 · 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.