Research map / Image generation / Autoregressive and latent generation
Hallucinated-IQA: No-Reference Image Quality Assessment via Adversarial Learning
Kwan-Yee Lin, Guanxiang Wang
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 |
|---|---|---|---|
| Pair distorted images with generated hallucinated references and their discrepancy maps (ablation) | NR-IQA regression on TID2008 · TID2008 | SROCC: 0.755 -> 0.859 | Improved |
| Add hallucinated-reference compensation (ablation) | NR-IQA regression on TID2008 · TID2008 | LCC: 0.800 -> 0.870 | Improved |
| Add quality-aware perceptual loss to the hallucination model (ablation) | BL+HCM on TID2008 · TID2008 | SROCC: 0.859 -> 0.864 | Improved |
| Add quality-aware perceptual loss to the hallucination model (ablation) | BL+HCM on TID2008 · TID2008 | LCC: 0.870 -> 0.887 | Improved |
| Add the original adversarial-learning mechanism (ablation) | BL+HCM+QPL on TID2008 · TID2008 | SROCC: 0.864 -> 0.894 | Improved |
| Add the original adversarial-learning mechanism (ablation) | BL+HCM+QPL on TID2008 · TID2008 | LCC: 0.887 -> 0.910 | Improved |
| Replace original adversarial learning with the proposed quality-aware adversarial mechanism (ablation) | BL+HCM+QPL+ADV on TID2008 · TID2008 | SROCC: 0.894 -> 0.910 | Improved |
| Replace original adversarial learning with the proposed quality-aware adversarial mechanism (ablation) | BL+HCM+QPL+ADV on TID2008 · TID2008 | LCC: 0.910 -> 0.918 | Improved |
| Add high-level semantic fusion between generator and regression features (ablation) | BL+HCM+QPL+QADV on TID2008 · TID2008 | SROCC: 0.910 -> 0.941 | Improved |
| Add high-level semantic fusion between generator and regression features (ablation) | BL+HCM+QPL+QADV on TID2008 · TID2008 | LCC: 0.918 -> 0.949 | Improved |
| Train the hallucination generator using TID2008 and TID2013 rather than LIVE alone (ablation) | Hallucinated-IQA trained on LIVE · LIVE | SROCC: 0.982 -> 0.982 | No clear effect |
Closest papers by meaning
- Adaptive Feature Selection for No-Reference Image Quality Assessment by Mitigating Semantic Noise Sensitivity (ICML 2024 · similarity 0.90)
- Backdoor Attacks Against No-Reference Image Quality Assessment Models via a Scalable Trigger (AAAI 2025 · similarity 0.87)
- Toward Generalized Image Quality Assessment: Relaxing the Perfect Reference Quality Assumption (CVPR 2025 · similarity 0.87)
- Augmenting Perceptual Super-Resolution via Image Quality Predictors (CVPR 2025 · similarity 0.87)
- Boosting Image Quality Assessment through Efficient Transformer Adaptation with Local Feature Enhancement (CVPR 2024 · similarity 0.87)
- Adaptive Image Quality Assessment via Teaching Large Multimodal Model to Compare (NeurIPS 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.
