Research map / Label-efficient and robust learning / Domain adaptation and generalization
Unsupervised Curriculum Domain Adaptation for No-Reference Video Quality Assessment
Pengfei Chen, Leida Li, Jinjian Wu, Weisheng Dong, Guangming Shi
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 curriculum-style second-stage subdomain adaptation (ablation) | UDA · LIVE VQA (source LBVD) | PLCC: 0.6749 -> 0.7797 | Improved |
| add curriculum-style second-stage subdomain adaptation (ablation) | UDA · CSIQ VQA (source LBVD) | PLCC: 0.7005 -> 0.8283 | Improved |
| add curriculum-style second-stage subdomain adaptation (ablation) | UDA · KoNViD-1k (source LBVD) | PLCC: 0.6981 -> 0.7909 | Improved |
| add curriculum-style second-stage subdomain adaptation (ablation) | UDA · LIVE-VQC (source LBVD) | PLCC: 0.7082 -> 0.7702 | Improved |
| predict rating distribution instead of scalar score (ablation) | UDA · LIVE VQA (source LBVD) | PLCC: 0.6749 -> 0.6833 | Improved |
| predict rating distribution instead of scalar score (ablation) | TA3N · LIVE VQA (source LBVD) | PLCC: 0.6917 -> 0.7005 | Improved |
| predict rating distribution instead of scalar score (ablation) | UDA · KoNViD-1k (source LBVD) | PLCC: 0.6981 -> 0.7146 | Improved |
| predict rating distribution instead of scalar score (ablation) | TA3N · KoNViD-1k (source LBVD) | PLCC: 0.7193 -> 0.7177 | Mixed |
| add unsupervised domain adaptation (stage one) to source-only training (ablation) | NoAdapt · KoNViD-1k (source LBVD) | PLCC: 0.6547 -> 0.6981 | Improved |
| use other VQA datasets as source domain, UCDA instead of UDA (ablation) | UDA · LIVE VQA target, averaged over source datasets | PLCC: 0.6428 -> 0.7663 | Improved |
| increase labeled target fraction in semi-supervised training (ablation) | UCDA · LBVD->KoNViD | PLCC: 0.8029 -> 0.8386 | Improved |
| increase split ratio eta from 0 up to 0.6 (ablation) | UCDA · five target databases | PLCC | Improved |
Closest papers by meaning
- MDS-VQA: Model-Informed Data Selection for Video Quality Assessment (CVPR 2026 · similarity 0.90)
- Generalizable Video Quality Assessment via Weak-to-Strong Learning (CVPR 2026 · similarity 0.89)
- PTM-VQA: Efficient Video Quality Assessment Leveraging Diverse PreTrained Models from the Wild (CVPR 2024 · similarity 0.89)
- Modular Blind Video Quality Assessment (CVPR 2024 · similarity 0.85)
- Bridging the Synthetic-to-Authentic Gap: Distortion-Guided Unsupervised Domain Adaptation for Blind Image Quality Assessment (CVPR 2024 · similarity 0.85)
- Image-to-video Adaptation with Outlier Modeling and Robust Self-learning (AAAI 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.
