No-Reference Image Quality Assessment on AGIQA-1K (test)
0.817SRCCCLIP-IQA+
Evaluation results
| Method | Links | ||
|---|---|---|---|
| CLIP-IQA+Opinion-Unaware=false2024.03 | 0.817 | 0.855 | |
| CONTRIQUEOpinion-Unaware=false2024.03 | 0.799 | 0.855 | |
| Re-IQAOpinion-Unaware=false2024.03 | 0.783 | 0.84 | |
| VisualQuality-R1Method Category=VLM-based, Training Protocol=Trained on KonIQ-10K2026.03 | 0.773 | 0.822 | |
| Q-Tacit†Method Category=VLM-based, Training Protocol=Multi-task training set in [25] for GRPO calibration2026.03 | 0.769 | 0.828 | |
| ARNIQAOpinion-Unaware=false2024.03 | 0.768 | 0.849 | |
| Q-InsightMethod Category=VLM-based, Training Protocol=Trained on KonIQ-10K2026.03 | 0.766 | 0.816 | |
| Q-TacitMethod Category=VLM-based, Training Protocol=Trained on KonIQ-10K2026.03 | 0.765 | 0.823 | |
| Q-Insight†Method Category=VLM-based, Training Protocol=Multi-task training set in [25] for GRPO calibration2026.03 | 0.764 | 0.811 | |
| GRepQOpinion-Unaware=false2024.03 | 0.74 | 0.797 | |
| Qwen-2.5-VL-7BMethod Category=VLM-based, Training Protocol=Trained on KonIQ-10K2026.03 | 0.739 | 0.813 | |
| QualiCLIPOpinion-Unaware=true2024.03 | 0.736 | 0.805 | |
| Q-AlignMethod Category=VLM-based, Training Protocol=Trained on KonIQ-10K2026.03 | 0.735 | 0.772 | |
| MDFSOpinion-Unaware=true2024.03 | 0.734 | 0.818 | |
| DeQA-ScoreMethod Category=VLM-based, Training Protocol=Trained on KonIQ-10K2026.03 | 0.729 | 0.809 | |
| CLIP-IQA+Method Category=Non-VLM Deep-learning, Training Protocol=Trained on KonIQ-10K2026.03 | 0.685 | 0.736 | |
| Compare2ScoreMethod Category=VLM-based, Training Protocol=Trained on KonIQ-10K2026.03 | 0.671 | 0.777 | |
| ARNIQA-OUOpinion-Unaware=true2024.03 | 0.667 | 0.786 | |
| NIMAMethod Category=Non-VLM Deep-learning, Training Protocol=Trained on KonIQ-10K2026.03 | 0.654 | 0.715 | |
| IL-NIQEOpinion-Unaware=true2024.03 | 0.645 | 0.757 | |
| CONTRIQUE-OUOpinion-Unaware=true2024.03 | 0.644 | 0.717 | |
| Re-IQA-OUOpinion-Unaware=true2024.03 | 0.644 | 0.764 | |
| ManIQAMethod Category=Non-VLM Deep-learning, Training Protocol=Trained on KonIQ-10K2026.03 | 0.636 | 0.723 | |
| MUSIQMethod Category=Non-VLM Deep-learning, Training Protocol=Trained on KonIQ-10K2026.03 | 0.63 | 0.722 | |
| NIQEOpinion-Unaware=true2024.03 | 0.623 | 0.721 | |
| NIQEMethod Category=Handcrafted, Training Protocol=Trained on KonIQ-10K2026.03 | 0.533 | 0.56 | |
| CLIP-IQAOpinion-Unaware=true2024.03 | 0.498 | 0.651 | |
| BRISQUEMethod Category=Handcrafted, Training Protocol=Trained on KonIQ-10K2026.03 | 0.497 | 0.541 | |
| CL-MIOpinion-Unaware=true2024.03 | 0.474 | 0.621 | |
| GRepQ-OUOpinion-Unaware=true2024.03 | 0.161 | 0.53 |