No-Reference Image Quality Assessment on CSIQ-IQA (test)
0.946SROCCDB-CNN
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DB-CNNModel Type=Supervised pretraining and supervised fine-tuning2021.10 | 0.946 | 0.959 | |
| CONTRIQUEModel Type=Unsupervised pretraining and Linear Regression2021.10 | 0.942 | 0.955 | |
| HyperIQAModel Type=Supervised pretraining and supervised fine-tuning2021.10 | 0.923 | 0.942 | |
| PQRModel Type=Supervised pretraining and supervised fine-tuning2021.10 | 0.872 | 0.901 | |
| Resnet-50Model Type=Supervised pretraining and Linear Regression2021.10 | 0.84 | 0.848 | |
| BIECONModel Type=Supervised pretraining and supervised fine-tuning2021.10 | 0.815 | 0.823 | |
| BRISQUEModel Type=Traditional/Handcrafted Features2021.10 | 0.746 | 0.829 | |
| HOSAModel Type=Codebook-based Features2021.10 | 0.741 | 0.823 | |
| CORNIAModel Type=Codebook-based Features2021.10 | 0.678 | 0.776 | |
| NIQEModel Type=Traditional/Handcrafted Features2021.10 | 0.627 | 0.712 |