Research map / Video / Action recognition
Procedural Generation of Videos to Train Deep Action Recognition Networks
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 |
|---|---|---|---|
| Add PHAV synthetic videos through Cool-TSN multi-task training | TSN · UCF-101 | accuracy: 93.6 -> 94.2 | Helped |
| Add PHAV synthetic videos through Cool-TSN multi-task training | TSN · HMDB-51 | accuracy: 66.6 -> 69.5 | Helped |
| Initialize fine-tuning with PHAV-trained weights | TSN · HMDB-51 | accuracy: 66.6 -> 68.9 | Helped |
| Initialize fine-tuning with PHAV-trained weights | TSN · UCF-101 | accuracy: 93.6 -> 94.1 | Helped |
| Use Cool-TSN with PHAV while reducing real training videos to 1% (ablation) | TSN · HMDB-51 | accuracy: 25.9 -> 27.7 | Helped |
| Use Cool-TSN with PHAV while reducing real training videos to 5% (ablation) | TSN · HMDB-51 | accuracy: 68.5 -> 71.5 | Helped |
| Use Cool-TSN with PHAV while reducing real training videos to 10% (ablation) | TSN · HMDB-51 | accuracy: 80.9 -> 84.4 | Helped |
| Use Cool-TSN with PHAV while reducing real training videos to 25% (ablation) | TSN · HMDB-51 | accuracy: 89.0 -> 90.4 | Helped |
| Use Cool-TSN with PHAV while reducing real training videos to 50% (ablation) | TSN · HMDB-51 | accuracy: 62.9 -> 65.8 | Helped |
| Use Cool-TSN with PHAV and all real training videos (ablation) | TSN · HMDB-51 | accuracy: 67.8 -> 70.1 | Helped |
Closest papers by meaning
- Learning Diverse Stochastic Human-Action Generators by Learning Smooth Latent Transitions (AAAI 2020 · similarity 0.88)
- Learning Video Representations of Human Motion From Synthetic Data (CVPR 2022 · similarity 0.88)
- Scaling Autoregressive Video Models (ICLR 2020 · similarity 0.88)
- HumanVid: Demystifying Training Data for Camera-controllable Human Image Animation (NeurIPS 2024 · similarity 0.86)
- Condensing Action Segmentation Datasets via Generative Network Inversion (CVPR 2025 · similarity 0.86)
- HumanDreamer: Generating Controllable Human-Motion Videos via Decoupled Generation (CVPR 2025 · 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.