Research map / Video / Action recognition
A Recurrent Transformer Network for Novel View Action Synthesis
Kara Marie Schatz, Erik Quintanilla, Shruti Vyas, Yogesh S Rawat
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 action transformation to basic model (ablation) | Basic model · NTU-RGB+D test | PSNR: 23.9 -> 24.7 | Improved |
| add action transformation to basic model (ablation) | Basic model · NTU-RGB+D test | SSIM: .940 -> .951 | Improved |
| add hierarchical transformation (ablation) | Basic + AC-Trans · NTU-RGB+D test | PSNR: 24.7 -> 26.7 | Improved |
| add hierarchical transformation (ablation) | Basic + AC-Trans · NTU-RGB+D test | SSIM: .951 -> .966 | Improved |
| add appearance transformation (recurrent, key-point guided) (ablation) | Basic + AC-Trans + HI-Trans · NTU-RGB+D test | PSNR: 26.7 -> 27.5 | Improved |
| add appearance transformation (recurrent, key-point guided) (ablation) | Basic + AC-Trans + HI-Trans · NTU-RGB+D test | SSIM: .966 -> .973 | Improved |
| use proposed method instead of VDNet, which uses target-view depth/skeleton | VDNet · NTU-RGB+D test | SSIM: .783 -> .973 | Improved |
| use proposed method instead of ResNet baseline | ResNet · NTU-RGB+D test | SSIM: .708 -> .973 | Improved |
| increase viewpoint gap from +/-45 to 90 degrees (v2<->v3 pairs) | Proposed · NTU-RGB+D test | SSIM: .975 -> .971 | Hurt |
Closest papers by meaning
- View-LSTM: Novel-View Video Synthesis Through View Decomposition (ICCV 2019 · similarity 0.90)
- Video Action Transformer Network (CVPR 2019 · similarity 0.88)
- RPAN: An End-To-End Recurrent Pose-Attention Network for Action Recognition in Videos (ICCV 2017 · similarity 0.88)
- Procedural Generation of Videos to Train Deep Action Recognition Networks (CVPR 2017 · similarity 0.87)
- DVANet: Disentangling View and Action Features for Multi-View Action Recognition (AAAI 2024 · similarity 0.87)
- Adapting Short-Term Transformers for Action Detection in Untrimmed Videos (CVPR 2024 · similarity 0.86)
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.
