Blind Image Quality Assessment on LIVEC
0.918SRCCDEFNet
Evaluation results
| Method | Links | ||
|---|---|---|---|
| DEFNet2025.07 | 0.918 | 0.897 | |
| DR.Expertspre-training=false2026.02 | 0.914 | 0.926 | |
| SHDIQANumber of trainable parameters=24M2026.05 | 0.909 | 0.922 | |
| CSFIQA2024.11 | 0.905 | 0.922 | |
| LIQE2025.07 | 0.904 | 0.91 | |
| LIQENumber of trainable parameters=151M2026.05 | 0.904 | 0.91 | |
| GLIANetNumber of trainable parameters=8M2026.05 | 0.899 | 0.913 | |
| QPTpre-training=true2026.02 | 0.895 | 0.914 | |
| QFM-IQMpre-training=false2026.02 | 0.891 | 0.913 | |
| QFM-IQM2024.11 | 0.891 | 0.913 | |
| LODApre-training=false2026.02 | 0.876 | 0.899 | |
| LoDa2024.11 | 0.876 | 0.899 | |
| LoDaNumber of trainable parameters=9M2026.05 | 0.876 | 0.899 | |
| DEIQTpre-training=false2026.02 | 0.875 | 0.894 | |
| QCNpre-training=false2026.02 | 0.875 | 0.893 | |
| DEIQT2024.11 | 0.875 | 0.894 | |
| DACNN2024.11 | 0.866 | 0.884 | |
| CDINet2024.11 | 0.865 | 0.88 | |
| CDINet2025.07 | 0.865 | 0.88 | |
| DPSF2025.07 | 0.865 | 0.882 | |
| LQMambapre-training=false2026.02 | 0.863 | 0.903 | |
| QMambaNumber of trainable parameters=50M2026.05 | 0.863 | 0.903 | |
| HyperIQApre-training=false2026.02 | 0.859 | 0.882 | |
| HyperIQA2024.11 | 0.859 | 0.882 | |
| TReS2024.11 | 0.859 | 0.882 | |
| QAL-IQA2025.07 | 0.859 | 0.875 | |
| HyperIQA2026.05 | 0.859 | 0.882 | |
| VCRNet2025.07 | 0.856 | 0.865 | |
| HyperIQA2025.07 | 0.855 | 0.878 | |
| UNIQUE2025.07 | 0.854 | 0.884 | |
| DBCNN2024.11 | 0.851 | 0.869 | |
| DBCNN2026.05 | 0.851 | 0.869 | |
| DPNet2025.07 | 0.849 | 0.864 | |
| TReSpre-training=false2026.02 | 0.846 | 0.877 | |
| TreS2025.07 | 0.846 | 0.877 | |
| TReS2026.05 | 0.846 | 0.877 | |
| CONRTIQUEpre-training=true2026.02 | 0.845 | 0.857 | |
| CONTRIQUE2025.07 | 0.845 | 0.857 | |
| DB-CNNpre-training=false2026.02 | 0.844 | 0.862 | |
| P2P-BM2024.11 | 0.844 | 0.842 | |
| P2P-BM2026.05 | 0.844 | 0.842 | |
| Re-IQA2024.11 | 0.84 | 0.854 | |
| Re-IQA2025.07 | 0.84 | 0.854 | |
| DAIQA2026.05 | 0.84 | 0.867 | |
| Re-IQANumber of trainable parameters=48M2026.05 | 0.84 | 0.854 | |
| MetaIQA2024.11 | 0.835 | 0.802 | |
| DBCNN2025.07 | 0.835 | 0.854 | |
| MetaIQA2026.05 | 0.835 | 0.802 | |
| CLIP-IQA+2024.11 | 0.805 | 0.832 | |
| MUSIQ2024.11 | 0.785 | 0.828 | |
| SynDR-IQATraining Dataset=KADID-10k2026.01 | 0.713 | 0.714 | |
| MUSIQpre-training=false2026.02 | 0.702 | 0.746 | |
| Q-AlignTraining Dataset=KADID-10k2026.01 | 0.702 | 0.744 | |
| MUSIQ2026.05 | 0.702 | 0.746 | |
| MEON2024.11 | 0.697 | 0.71 | |
| DGQATraining Dataset=KADID-10k2026.01 | 0.696 | 0.69 | |
| WaDIQaM-NRpre-training=false2026.02 | 0.692 | 0.73 | |
| BRISQUE2024.11 | 0.629 | 0.629 | |
| FreqAlignTraining Dataset=KADID-10k2026.01 | 0.618 | 0.588 | |
| BIECON2024.11 | 0.613 | 0.613 | |
| BRISQUEpre-training=false2026.02 | 0.601 | 0.621 | |
| StyleAMTraining Dataset=KADID-10k2026.01 | 0.584 | 0.561 | |
| KGANetTraining Dataset=KADID-10k2026.01 | 0.575 | — | |
| DBCNNTraining Dataset=KADID-10k2026.01 | 0.572 | 0.589 | |
| VCRNetTraining Dataset=KADID-10k2026.01 | 0.561 | 0.548 | |
| MUSIQTraining Dataset=KADID-10k2026.01 | 0.517 | 0.524 | |
| CLIPIQA+Training Dataset=KADID-10k2026.01 | 0.512 | 0.543 | |
| ILNIQE2024.11 | 0.508 | 0.508 | |
| ILNIQE2026.05 | 0.508 | 0.508 | |
| DANNTraining Dataset=KADID-10k2026.01 | 0.499 | 0.484 | |
| RankIQATraining Dataset=KADID-10k2026.01 | 0.491 | 0.495 | |
| HyperIQATraining Dataset=KADID-10k2026.01 | 0.49 | 0.487 | |
| ILNIQE2025.07 | 0.469 | 0.518 | |
| ILNIQEpre-training=false2026.02 | 0.453 | 0.511 | |
| RankDATraining Dataset=KADID-10k2026.01 | 0.451 | 0.455 | |
| NIQE2025.07 | 0.446 | 0.507 | |
| UCDATraining Dataset=KADID-10k2026.01 | 0.382 | 0.358 | |
| Ma192025.07 | 0.336 | 0.405 | |
| dipIQ2025.07 | 0.187 | 0.29 |