Image Quality Assessment on LIVE-Wild
0.909PLCCQ-Hawkeye
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Q-HawkeyeTraining Dataset=KonIQ2026.01 | 0.909 | 0.88 | |
| Q-HawkeyeMethodology=MLLM-Based Methods2026.01 | 0.909 | 0.88 | |
| DeQA-ScoreTraining Dataset=KonIQ, SPAQ2026.01 | 0.902 | 0.888 | |
| DeQA-ScoreTraining Dataset=KonIQ, SPAQ, KADID2026.01 | 0.9 | 0.887 | |
| RALICategory=Non-MLLM Deep-learning, Training Dataset=KonIQ2025.10 | 0.896 | 0.876 | |
| Q-insightTraining Dataset=KonIQ, KADIS2026.01 | 0.893 | 0.865 | |
| Q-InsightCategory=MLLM-based, Training Dataset=KonIQ2025.10 | 0.893 | 0.865 | |
| DeQA-ScoreVenue=CVPR’25, Methodology=MLLM-Based Methods2026.01 | 0.892 | 0.879 | |
| DeQACategory=VLM (w/o & w/ reasoning), Trained on KonIQ=true2026.01 | 0.892 | 0.879 | |
| DeQACategory=MLLM-based, Training Dataset=KonIQ2025.10 | 0.892 | 0.879 | |
| Zoom-IQACategory=VLM (w/o & w/ reasoning), Trained on KonIQ=true2026.01 | 0.887 | 0.87 | |
| DeQA-ScoreTraining Dataset=KonIQ, SPAQ, KADID, PIPAL2026.01 | 0.877 | 0.857 | |
| VisualQuality-R1Category=MLLM-based, Training Dataset=KonIQ2025.10 | 0.874 | 0.849 | |
| Q-insightVenue=NeurIPS’25, Methodology=MLLM-Based Methods2026.01 | 0.87 | 0.839 | |
| Q-InsightCategory=VLM (w/o & w/ reasoning), Trained on KonIQ=true2026.01 | 0.87 | 0.839 | |
| VisualQuality-R1Training Dataset=KADID, SPAQ2026.01 | 0.856 | 0.827 | |
| VisualQuality-R1Venue=NeurIPS’25, Methodology=MLLM-Based Methods2026.01 | 0.856 | 0.827 | |
| VisualQuality-R1Category=VLM (w/o & w/ reasoning), Trained on KonIQ=true2026.01 | 0.856 | 0.827 | |
| Q-AlignVenue=ICML’24, Methodology=MLLM-Based Methods2026.01 | 0.853 | 0.86 | |
| Q-AlignCategory=VLM (w/o & w/ reasoning), Trained on KonIQ=true2026.01 | 0.853 | 0.86 | |
| Q-AlignCategory=MLLM-based, Training Dataset=KonIQ2025.10 | 0.853 | 0.86 | |
| ManIQAVenue=CVPR’22, Methodology=Non-MLLM Deep-Learning2026.01 | 0.849 | 0.832 | |
| ManIQACategory=Non-VLM, Trained on KonIQ=true2026.01 | 0.849 | 0.832 | |
| LIQECategory=Non-MLLM Deep-learning, Training Dataset=KonIQ2025.10 | 0.842 | 0.865 | |
| CLIP-IQA+Category=VLM (w/o & w/ reasoning), Trained on KonIQ=true2026.01 | 0.832 | 0.805 | |
| CLIP-IQA+Category=Non-MLLM Deep-learning, Training Dataset=KonIQ2025.10 | 0.832 | 0.832 | |
| NIMAVenue=TIP’18, Methodology=Non-MLLM Deep-Learning2026.01 | 0.814 | 0.771 | |
| NIMACategory=Non-VLM, Trained on KonIQ=true2026.01 | 0.814 | 0.771 | |
| CLIP-IQA+Venue=AAAI’23, Methodology=Non-MLLM Deep-Learning2026.01 | 0.805 | 0.752 | |
| ManIQACategory=Non-MLLM Deep-learning, Training Dataset=KonIQ2025.10 | 0.805 | 0.849 | |
| MUSIQVenue=ICCV’21, Methodology=Non-MLLM Deep-Learning2026.01 | 0.789 | 0.83 | |
| MUSIQCategory=Non-VLM, Trained on KonIQ=true2026.01 | 0.789 | 0.83 | |
| MUSIQCategory=Non-MLLM Deep-learning, Training Dataset=KonIQ2025.10 | 0.789 | 0.83 | |
| Compare2ScoreVenue=NIPS’24, Methodology=MLLM-Based Methods2026.01 | 0.786 | 0.772 | |
| C2ScoreCategory=VLM (w/o & w/ reasoning), Trained on KonIQ=true2026.01 | 0.786 | 0.772 | |
| C2ScoreCategory=MLLM-based, Training Dataset=KonIQ2025.10 | 0.786 | 0.772 | |
| DBCNNVenue=ICSIPA’19, Methodology=Non-MLLM Deep-Learning2026.01 | 0.773 | 0.735 | |
| DBCNNCategory=Non-VLM, Trained on KonIQ=true2026.01 | 0.773 | 0.755 | |
| HyperIQAVenue=CVPR’20, Methodology=Non-MLLM Deep-Learning2026.01 | 0.772 | 0.749 | |
| HyperIQACategory=Non-VLM, Trained on KonIQ=true2026.01 | 0.772 | 0.749 | |
| Qwen-SFTVenue=arXiv’25, Methodology=MLLM-Based Methods2026.01 | 0.734 | 0.728 | |
| NIQEVenue=SPL’12, Methodology=Handcrafted2026.01 | 0.493 | 0.449 | |
| NIQECategory=Handcrafted, Trained on KonIQ=false2026.01 | 0.493 | 0.449 | |
| NIQECategory=Handcrafted, Training Dataset=N/A2025.10 | 0.493 | 0.449 | |
| BRISQUEVenue=TIP’12, Methodology=Handcrafted2026.01 | 0.361 | 0.313 | |
| BRISQUECategory=Handcrafted, Trained on KonIQ=false2026.01 | 0.361 | 0.313 | |
| BRISQUECategory=Handcrafted, Training Dataset=N/A2025.10 | 0.361 | 0.313 |