Video Quality Assessment on LIVE-Qualcomm (test)
0.8387SRCCPVQ
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| PVQTraining Setting=Mixed Databases2021.08 | 0.8387 | 0.8617 | |
| Li et al.Model Type=Supervised Pretraining2022.06 | 0.833 | 0.837 | |
| Li et al.Model Type=Supervised Pretraining2022.06 | 0.833 | 0.837 | |
| FasterVQAResolution Range=1080P, Efficiency=4X than FAST-VQA2022.10 | 0.826 | 0.844 | |
| FAST-VQAGroup=Ours, Protocol=Finetune2022.07 | 0.819 | 0.851 | |
| FAST-VQAResolution Range=1080P2022.10 | 0.819 | 0.851 | |
| BVQA-TCSVT-2022Group=Existing Fixed Deep, Resolution Range=1080P2022.10 | 0.817 | 0.828 | |
| MDTVSFATraining Setting=Mixed Databases2021.08 | 0.8136 | 0.8291 | |
| CNN+TLVQMGroup=Ensemble C+D, Protocol=Finetune2022.07 | 0.81 | 0.833 | |
| CNN+TLVQMGroup=Ensemble C+D, Resolution Range=1080P2022.10 | 0.81 | 0.833 | |
| Proposed-LSTraining Setting=Mixed Databases, Scaling=Linear Re-scaling (LS)2021.08 | 0.8041 | 0.8253 | |
| FAST-VQA-MGroup=Ours, Protocol=Finetune2022.07 | 0.804 | 0.838 | |
| FAST-VQA-MResolution Range=1080P2022.10 | 0.804 | 0.838 | |
| GSTVQAModel Type=Supervised Pretraining2022.06 | 0.801 | 0.825 | |
| GSTVQAModel Type=Supervised Pretraining2022.06 | 0.801 | 0.825 | |
| GST-VQAGroup=Existing Fixed Deep, Protocol=Finetune2022.07 | 0.801 | 0.825 | |
| GST-VQAGroup=Existing Fixed Deep, Resolution Range=1080P2022.10 | 0.801 | 0.825 | |
| CONVIQTModel Type=Unsupervised Pretraining2022.06 | 0.797 | 0.802 | |
| CONVIQTModel Type=Unsupervised Pretraining2022.06 | 0.797 | 0.802 | |
| Full-res Swin-T featuresGroup=Baseline Features, Protocol=Finetune2022.07 | 0.788 | 0.803 | |
| Full-res Swin-TResolution Range=1080P, Features=Full-resolution Swin-T2022.10 | 0.788 | 0.803 | |
| TLVQMModel Type=Traditional/Handcrafted Features2022.06 | 0.785 | 0.815 | |
| TLVQMModel Type=Traditional/Handcrafted Features2022.06 | 0.785 | 0.815 | |
| TLVQMGroup=Existing Classical, Protocol=Finetune2022.07 | 0.77 | 0.81 | |
| TLVQMGroup=Existing Classical, Resolution Range=1080P2022.10 | 0.77 | 0.81 | |
| CONTRIQUEModel Type=Unsupervised Pretraining2022.06 | 0.765 | 0.777 | |
| CONTRIQUEModel Type=Unsupervised Pretraining2022.06 | 0.765 | 0.777 | |
| FAST-VQA w/o VQ-representationsGroup=Ours, Protocol=Finetune2022.07 | 0.756 | 0.778 | |
| VSFAGroup=Existing Fixed Deep, Protocol=Finetune2022.07 | 0.737 | 0.779 | |
| VSFAGroup=Existing Fixed Deep, Resolution Range=1080P2022.10 | 0.737 | 0.732 | |
| VSFAModel Type=Supervised Pretraining2022.06 | 0.708 | 0.774 | |
| VSFAModel Type=Supervised Pretraining2022.06 | 0.708 | 0.774 | |
| VIDEVALModel Type=Traditional/Handcrafted Features2022.06 | 0.67 | 0.705 | |
| VIDEVALModel Type=Traditional/Handcrafted Features2022.06 | 0.67 | 0.705 | |
| RAPIQUEModel Type=Traditional/Handcrafted Features2022.06 | 0.665 | 0.691 | |
| RAPIQUEModel Type=Traditional/Handcrafted Features2022.06 | 0.665 | 0.691 | |
| Resnet-50Model Type=Supervised Pretraining, Pretrained=ImageNet2022.06 | 0.65 | 0.691 | |
| Resnet-50Model Type=Supervised Pretraining2022.06 | 0.65 | 0.691 | |
| VBLIINDSModel Type=Traditional/Handcrafted Features2022.06 | 0.57 | 0.626 | |
| VBLIINDSModel Type=Traditional/Handcrafted Features2022.06 | 0.57 | 0.626 | |
| BRISQUEModel Type=Traditional/Handcrafted Features2022.06 | 0.552 | 0.598 | |
| BRISQUEModel Type=Traditional/Handcrafted Features2022.06 | 0.552 | 0.598 |