Research map / Label-efficient and robust learning / Domain adaptation and generalization
VideoDG: Generalizing Temporal Relations in Videos to Novel Domains
Zhiyu Yao, Yunbo Wang, Jianmin Wang, Philip S. Yu, Mingsheng Long
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 RADA adversarial augmentation to APN | APN · UCF->HMDB | accuracy (%): 54.3 -> 59.1 | Improved |
| add RADA adversarial augmentation to APN | APN · HMDB->UCF | accuracy (%): 71.4 -> 74.9 | Improved |
| apply original ADA (semantic) to NL-I3D | NL-I3D · UCF->HMDB | accuracy (%): 54.0 -> 53.5 | Hurt |
| use pixel-space instead of semantic-space constraint (ADA_pixel vs RADA) (ablation) | APN · UCF->HMDB | accuracy (%): RADA outperforms ADA_pixel by 2.2% | Improved |
| add RADA instead of M-ADA | APN · UCF->HMDB | accuracy (%): 55.6 -> 59.1 | Improved |
| add random hard augmentation to APN | APN · UCF->HMDB | accuracy (%): 54.3 -> 56.1 | Improved |
| use Jigsaw puzzle instead of RADA | APN · UCF->HMDB | accuracy (%): 59.1 -> 55.2 | Hurt |
| generate adversarial data with level II and III features instead of level I only (ablation) | APN + RADA · UCF->HMDB | accuracy (%): 53.6 -> 59.1 | Improved |
| add level I features to final II+III (ablation) | APN + RADA · UCF->HMDB | accuracy (%): 59.1 -> 58.6 | Hurt |
| set T_max to 5 instead of 1 (ablation) | RADA · UCF->HMDB | accuracy (%): 58.1 -> 59.1 | Improved |
| set T_max to 15 instead of 5 (ablation) | RADA · UCF->HMDB | accuracy (%): 59.1 -> 58.1 | Hurt |
| enable robustness loss (lambda=1 vs 0) (ablation) | RADA · UCF->HMDB | accuracy (%) | Improved |
Closest papers by meaning
- VUDG: A Dataset for Video Understanding Domain Generalization (ICLR 2026 · similarity 0.90)
- Diversifying Spatial-Temporal Perception for Video Domain Generalization (NeurIPS 2023 · similarity 0.89)
- Episodic Training for Domain Generalization (ICCV 2019 · similarity 0.88)
- Deeper, Broader and Artier Domain Generalization (ICCV 2017 · similarity 0.87)
- Exploiting Feature and Class Relationships in Video Categorization with Regularized Deep Neural Networks (TPAMI 2017 · similarity 0.87)
- PracticalDG: Perturbation Distillation on Vision-Language Models for Hybrid Domain Generalization (CVPR 2024 · 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.
