Blind Image Quality Assessment on LIVE-W (test)
0.892PLCCDeQA
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DeQAModel category=SFT-based and RL-based MLLMs, Input Condition=Image-conditioned2026.01 | 0.892 | 0.879 | |
| DeQACategory=SFT-based and RL-based MLLMs, Condition=Image-conditioned2025.12 | 0.892 | 0.879 | |
| Chain-of-ThoughtModel category=SFT-based and RL-based MLLMs, Input Condition=Image-conditioned2026.01 | 0.878 | 0.851 | |
| OursCategory=SFT-based and RL-based MLLMs, Condition=Image-conditioned2025.12 | 0.877 | 0.849 | |
| Self-ConsistencyModel category=SFT-based and RL-based MLLMs, Input Condition=Image-conditioned2026.01 | 0.874 | 0.843 | |
| Autoencoder-likeModel category=SFT-based and RL-based MLLMs, Input Condition=Image-conditioned2026.01 | 0.873 | 0.85 | |
| Q-Insight-ScoreModel category=SFT-based and RL-based MLLMs, Input Condition=Image-conditioned2026.01 | 0.87 | 0.839 | |
| Q-Insight-ScoreCategory=SFT-based and RL-based MLLMs, Condition=Image-conditioned2025.12 | 0.87 | 0.839 | |
| Q-AlignModel category=SFT-based and RL-based MLLMs, Input Condition=Image-conditioned2026.01 | 0.853 | 0.86 | |
| Q-AlignCategory=SFT-based and RL-based MLLMs, Condition=Image-conditioned2025.12 | 0.853 | 0.86 | |
| MANIQAModel category=Deep-Learning Models, Input Condition=Image-conditioned2026.01 | 0.849 | 0.832 | |
| MANIQACategory=Deep-Learning Models, Condition=Image-conditioned2025.12 | 0.849 | 0.832 | |
| Self-ConsistencyInput Condition=Text-Only2026.01 | 0.838 | 0.815 | |
| CLIP-IQA+Model category=Deep-Learning Models, Input Condition=Image-conditioned2026.01 | 0.832 | 0.805 | |
| CLIP-IQA+Category=Deep-Learning Models, Condition=Image-conditioned2025.12 | 0.832 | 0.805 | |
| Self-ConsistencyInput Condition=Text-Only, Text Filtering=Score-related words removed2026.01 | 0.829 | 0.812 | |
| NIMAModel category=Deep-Learning Models, Input Condition=Image-conditioned2026.01 | 0.814 | 0.771 | |
| NIMACategory=Deep-Learning Models, Condition=Image-conditioned2025.12 | 0.814 | 0.771 | |
| OursCategory=Caption-Only Conditions, Condition=Caption-Only2025.12 | 0.791 | 0.798 | |
| MUSIQModel category=Deep-Learning Models, Input Condition=Image-conditioned2026.01 | 0.789 | 0.83 | |
| MUSIQCategory=Deep-Learning Models, Condition=Image-conditioned2025.12 | 0.789 | 0.83 | |
| C2ScoreModel category=SFT-based and RL-based MLLMs, Input Condition=Image-conditioned2026.01 | 0.786 | 0.772 | |
| C2ScoreCategory=SFT-based and RL-based MLLMs, Condition=Image-conditioned2025.12 | 0.786 | 0.772 | |
| Chain-of-ThoughtInput Condition=Text-Only2026.01 | 0.779 | 0.776 | |
| Q-Insight-ScoreInput Condition=Text-Only2026.01 | 0.778 | 0.776 | |
| DBCNNModel category=Deep-Learning Models, Input Condition=Image-conditioned2026.01 | 0.773 | 0.73 | |
| DBCNNCategory=Deep-Learning Models, Condition=Image-conditioned2025.12 | 0.773 | 0.73 | |
| Q-Insight-ScoreInput Condition=Text-Only, Text Filtering=Score-related words removed2026.01 | 0.772 | 0.771 | |
| Chain-of-ThoughtInput Condition=Text-Only, Text Filtering=Score-related words removed2026.01 | 0.772 | 0.771 | |
| Q-Insight-ScoreCategory=Caption-Only Conditions, Condition=Caption-Only2025.12 | 0.769 | 0.757 | |
| Autoencoder-likeInput Condition=Text-Only2026.01 | 0.761 | 0.767 | |
| Autoencoder-likeInput Condition=Text-Only, Text Filtering=Score-related words removed2026.01 | 0.758 | 0.763 | |
| NIQECategory=Hand-Crafted Models, Condition=Image-conditioned2025.12 | 0.493 | 0.449 | |
| BRISQUECategory=Hand-Crafted Models, Condition=Image-conditioned2025.12 | 0.361 | 0.313 |