Research map / Image generation / Image restoration
Quality-Aware Pre-Trained Models for Blind Image Quality Assessment
Kai Zhao, Kun Yuan, Ming Sun, Mading Li, Xing Wen
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 |
|---|---|---|---|
| Use QPT pre-training instead of standard ImageNet-supervised pre-training | ResNet50 BIQA model · BID | SRCC: 0.8423 -> 0.8875 | Improved |
| Use QPT pre-training instead of standard ImageNet-supervised pre-training | ResNet50 BIQA model · CLIVE | SRCC: 0.8527 -> 0.8947 | Improved |
| Replace HyperIQA pre-trained weights with QPT weights | HyperIQA · BID | SRCC: 0.8665 -> 0.8982 | Improved |
| Replace TReS pre-trained weights with QPT weights | TReS · CLIVE | SRCC: 0.8436 -> 0.8666 | Improved |
| Increase QPT pre-training data from 20% to 100% of ImageNet (ablation) | QPT ResNet50 · BID | SRCC: 0.8311 -> 0.8875 | Improved |
| Increase QPT pre-training data from 20% to 100% of ImageNet (ablation) | QPT ResNet50 · CLIVE | SRCC: 0.8115 -> 0.8947 | Improved |
| Use ResNet-50 rather than ResNet-18 encoder (ablation) | QPT · BID | SRCC: 0.8679 -> 0.8875 | Improved |
| Add inter-sample content-based negatives (ablation) | QPT contrastive loss with intra-sample negatives · BID | SRCC: 0.8399 -> 0.8875 | Improved |
| Add intra-sample degradation-based negatives (ablation) | QPT contrastive loss with inter-sample negatives · BID | SRCC: 0.8702 -> 0.8875 | Improved |
| Remove both inter- and intra-sample negative pairs (ablation) | QPT contrastive loss · BID | SRCC: 0.8875 -> 0.7078 | Hurt |
| Add two-order degradation (ablation) | QPT degradation space with skip and shuffle · BID | SRCC: 0.8828 -> 0.8875 | Improved |
| Add shuffle to degradation composition (ablation) | QPT degradation space with skip · BID | SRCC: 0.8745 -> 0.8875 | Improved |
Closest papers by meaning
- Integrating Global Context Contrast and Local Sensitivity for Blind Image Quality Assessment (ICML 2024 · similarity 0.92)
- Feature Denoising Diffusion Model for Blind Image Quality Assessment (AAAI 2025 · similarity 0.91)
- Semi-Supervised Blind Image Quality Assessment through Knowledge Distillation and Incremental Learning (AAAI 2024 · similarity 0.90)
- Distilling Spatially-Heterogeneous Distortion Perception for Blind Image Quality Assessment (CVPR 2025 · similarity 0.89)
- Bridging the Synthetic-to-Authentic Gap: Distortion-Guided Unsupervised Domain Adaptation for Blind Image Quality Assessment (CVPR 2024 · similarity 0.89)
- Towards Syn-to-Real IQA: A Novel Perspective on Reshaping Synthetic Data Distributions (NeurIPS 2025 · similarity 0.88)
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.
