Blind Image Quality Assessment on PIQ Overall 1.0 (test)
0.85SRCCDN-PIQA
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| DN-PIQABackbone=Swin Transformer-B, Training Data=LSVQ and GFIQA2024.05 | 0.85 | 0.86 | 0.68 | 0.53 | |
| PICNIQ2024.05 | 0.81 | 0.81 | 0.61 | 0.72 | |
| LIQE2024.05 | 0.81 | 0.8 | 0.62 | 0.64 | |
| FHIQA2024.05 | 0.78 | 0.78 | 0.59 | 1.12 | |
| MUSIQTraining Data=KonI0-10k2024.05 | 0.76 | 0.75 | 0.57 | 0.95 | |
| SEM-HyperIQA2024.05 | 0.75 | 0.75 | 0.56 | 1.03 | |
| MUSIQTraining Data=PaQ-2-PiQ2024.05 | 0.74 | 0.74 | 0.54 | 1.09 | |
| HyperIQA2024.05 | 0.74 | 0.74 | 0.55 | 0.99 | |
| SEM-HyperIQA-CO2024.05 | 0.74 | 0.74 | 0.55 | 1.04 | |
| StairIQA2024.05 | 0.72 | 0.75 | 0.55 | 0.7 | |
| DB-CNNModel Variant=CIVE C2024.05 | 0.59 | 0.64 | 0.43 | 1.04 |