Research map / 3D vision / Human pose estimation
Joint Dynamic Pose Image and Space Time Reversal for Human Action Recognition from Videos
Mengyuan Liu, Fanyang Meng, Chen Chen, Songtao Wu
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 attention maps from joint estimation maps to DTI-f (ablation) | DTI-f · NTU | accuracy: 70.60% -> 73.83% | Improved |
| add attention maps to DTI-t (ablation) | DTI-t · NTU | accuracy: 70.13% -> 79.86% | Improved |
| add attention maps to combined DTIs (ablation) | DTIs · NTU | accuracy: 85.39% -> 88.02% | Improved |
| combine three DTI views instead of a single view (ablation) | DTIs · NTU | accuracy: 77.49% -> 85.39% | Improved |
| late-fuse DPI with att-DTIs (ablation) | att-DTIs · NTU | accuracy: 88.02% -> 90.23% | Improved |
| late-fuse DPI with att-DTIs (ablation) | att-DTIs · UTD-MHAD | accuracy: 85.81% -> 88.37% | Improved |
| late-fuse DPI with att-DTIs (ablation) | att-DTIs · Penn-Action | accuracy: 94.25% -> 95.86% | Improved |
| late-fuse DPI with att-DTI-f (ablation) | att-DTI-f · NTU | accuracy: 73.83% -> 84.12% | Improved |
| use side view (s) instead of front view for DTI (ablation) | DTI · NTU | accuracy: 70.60% -> 77.49% | Improved |
| use side view (s) instead of front view for DTI (ablation) | DTI · UTD-MHAD | accuracy: 68.28% -> 65.16% | Hurt |
| increase channel number K from 3 to 14 (ablation) | DTI-f · Penn-Action | accuracy: 77.34% -> 83.50% | Improved |
| increase channel number K from 14 to 28 (ablation) | DTI-f · Penn-Action | accuracy | Hurt |
Closest papers by meaning
- Video-Based Human Pose Regression via Decoupled Space-Time Aggregation (CVPR 2024 · similarity 0.86)
- MMNet: A Model-Based Multimodal Network for Human Action Recognition in RGB-D Videos (TPAMI 2022 · similarity 0.86)
- Dynamic Kernel Distillation for Efficient Pose Estimation in Videos (ICCV 2019 · similarity 0.86)
- UCF-STAR: A Large Scale Still Image Dataset for Understanding Human Actions (AAAI 2020 · similarity 0.86)
- Optimizing Human Pose Estimation Through Focused Human and Joint Regions (AAAI 2025 · similarity 0.85)
- Dynamic Camera Poses and Where to Find Them (CVPR 2025 · similarity 0.85)
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.
