Research map / Video / Action recognition
PRISM: Learning a Shared Primitive Space for Transferable Skeleton Action Representation
Di Yang, Yaohui Wang, Shuai Shao, François Brémond, Jiangtao Wang
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 motion primitive representation (ablation) | PRISM · Toyota Smarthome Trimmed CV2 | accuracy (%): 66.4 -> 72.5 | Improved |
| add motion primitive representation (ablation) | PRISM · Toyota Smarthome Untrimmed CV | mAP (%): 23.6 -> 28.1 | Improved |
| add orthogonality and sparsity dictionary constraints (ablation) | PRISM · Toyota Smarthome Trimmed CV2 | accuracy (%): 70.3 -> 72.5 | Improved |
| add temporal feature fusion (TFF) (ablation) | PRISM · Toyota Smarthome Untrimmed CV | mAP (%): 24.1 -> 28.1 | Improved |
| add TFF (ablation) | PRISM · Toyota Smarthome Trimmed CV2 | accuracy (%): 71.7 -> 72.5 | Improved |
| add primitive coefficients to context features (ablation) | PRISM · Toyota Smarthome Trimmed CV2 | accuracy (%): 65.2 -> 72.5 | Improved |
| add context features to primitive coefficients (ablation) | PRISM · Toyota Smarthome Untrimmed CV | mAP (%): 11.8 -> 28.1 | Improved |
| use element-wise addition instead of attention fusion (ablation) | PRISM · Toyota Smarthome Trimmed CV2 | accuracy (%): 66.7 -> 72.5 | Improved |
| use element-wise addition instead of concatenation (ablation) | PRISM · Toyota Smarthome Trimmed CV2 | accuracy (%): 72.0 -> 72.5 | No clear effect |
| add orthogonality loss to decomposition objective (ablation) | PRISM · Mixamo motion transfer | Mean Square Error: 1.35 -> 0.83 | Improved |
| add sparsity loss (ablation) | PRISM · Mixamo motion transfer | Mean Square Error: 0.83 -> 0.82 | No clear effect |
| add physics-based loss (ablation) | PRISM · Mixamo motion transfer | Mean Square Error: 0.82 -> 0.80 | Improved |
Closest papers by meaning
- Revealing Key Details to See Differences: A Novel Prototypical Perspective for Skeleton-based Action Recognition (CVPR 2025 · similarity 0.89)
- Heterogeneous Skeleton-Based Action Representation Learning (CVPR 2025 · similarity 0.89)
- Neuron: Learning Context-Aware Evolving Representations for Zero-Shot Skeleton Action Recognition (CVPR 2025 · similarity 0.88)
- SAMoR: Motion Modelling for Articulated Objects of Any Skeleton and Topology (NeurIPS 2026 · similarity 0.87)
- Recovering Complete Actions for Cross-dataset Skeleton Action Recognition (NeurIPS 2024 · similarity 0.87)
- Part-aware Unified Representation of Language and Skeleton for Zero-shot Action Recognition (CVPR 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.
