Research map / Video / Action recognition
Histogram of Oriented Principal Components for Cross-View Action Recognition
Hossein Rahmani, Arif Mahmood, Du Q. Huynh, Ajmal Saeed Mian
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 |
|---|---|---|---|
| combine STK-D with Local HOPC (ablation) | Local HOPC · Northwestern-UCLA Multiview | accuracy (%): 71.9 -> 80.0 | Improved |
| combine STK-D with Local HOPC (ablation) | Local HOPC · UWA3D Multiview Activity II | accuracy (%): 44.0 -> 52.2 | Improved |
| use subject-height-relative spatial scale instead of constant scale (ablation) | Local HOPC+STK-D · Northwestern-UCLA Multiview | accuracy (%): 77.9 -> 80.0 | Improved |
| use subject-height-relative spatial scale instead of constant scale (ablation) | Local HOPC+STK-D · UWA3D Multiview Activity II | accuracy (%): 48.0 -> 52.2 | Improved |
| use subject-height scale instead of automatic scale selection (ablation) | Local HOPC+STK-D · Northwestern-UCLA Multiview | accuracy (%): 79.5 -> 80.0 | Improved |
| use subject-height scale instead of automatic scale selection (ablation) | Local HOPC+STK-D · UWA3D Multiview Activity II | accuracy (%): 50.9 -> 52.2 | Improved |
| add automatic temporal scale selection instead of constant tau=2 (ablation) | Local HOPC+STK-D · Northwestern-UCLA Multiview | accuracy (%): 78.3 -> 80.0 | Improved |
| add automatic temporal scale selection instead of constant tau=2 (ablation) | Local HOPC+STK-D · UWA3D Multiview Activity II | accuracy (%): 49.2 -> 52.2 | Improved |
| select top 98% discriminative codewords by F-score out of 1500 | HOPC bag-of-words · action recognition | accuracy (%): up to 1.5% improved accuracy | Improved |
| vary number of STKs from 100 to 1000 (ablation) | Local HOPC+STK-D · average accuracy | accuracy (%): best at 400 | Mixed |
| vary eigenratio threshold for STK detection between 1.1 and 1.5 (ablation) | Local HOPC+STK-D · average accuracy | accuracy (%): stable for 1.1 to 1.5 | No clear effect |
| increase local eigenratio threshold above 1.5 (ablation) | Local HOPC · average accuracy | accuracy (%): decreases when threshold > 1.5 | Hurt |
Closest papers by meaning
- 3DInAction: Understanding Human Actions in 3D Point Clouds (CVPR 2024 · similarity 0.85)
- Multi-view Action Recognition using Cross-view Video Prediction (ECCV 2020 · similarity 0.84)
- Cross-Modal Learning with 3D Deformable Attention for Action Recognition (ICCV 2023 · similarity 0.83)
- 3DV: 3D Dynamic Voxel for Action Recognition in Depth Video (CVPR 2020 · similarity 0.83)
- Adapting Pre-trained 3D Models for Point Cloud Video Understanding via Cross-frame Spatio-temporal Perception (CVPR 2025 · similarity 0.82)
- DVANet: Disentangling View and Action Features for Multi-View Action Recognition (AAAI 2024 · 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.
