Image Quality Assessment on Deblur
89.1SRCCQ-DeepSight
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Q-DeepSight2026.04 | 89.1 | 89.7 | |
| Tool-IQATraining dataset=KADID-10K, Method category=VLM-based2026.06 | 84.8 | 87.8 | |
| VisualQuality-R12026.04 | 84.5 | 87.9 | |
| Q-Insight2026.04 | 83.1 | 85.7 | |
| QInsightTraining dataset=KADID-10K, Method category=VLM-based2026.06 | 83.1 | 85.7 | |
| Qwen2.5-VL-7BTraining dataset=KADID-10K, Method category=VLM-based2026.06 | 82 | 85.2 | |
| VisualQuality-R1Training dataset=KADID-10K, Method category=VLM-based2026.06 | 81.1 | 85.9 | |
| Qwen3-VL-8B-InsTraining dataset=KADID-10K, Method category=VLM-based2026.06 | 81 | 84.2 | |
| LIQE2026.04 | 79.7 | 71.2 | |
| LIQETraining dataset=KADID-10K, Method category=VLM-based2026.06 | 79.7 | 71.2 | |
| DeQA-Score2026.04 | 78.5 | 83.8 | |
| DeQA-ScoreTraining dataset=KADID-10K, Method category=VLM-based2026.06 | 78.5 | 83.8 | |
| Q-Align2026.04 | 76.1 | 80.2 | |
| Q-AlignTraining dataset=KADID-10K, Method category=VLM-based2026.06 | 76.1 | 80.2 | |
| DEIQTTraining dataset=KADID-10K, Method category=Discriminative Deep-Learning-based2026.06 | 73.4 | 75.9 | |
| MANIQATraining dataset=KADID-10K, Method category=Discriminative Deep-Learning-based2026.06 | 72.6 | 72.8 | |
| UNIQUETraining dataset=KADID-10K, Method category=Discriminative Deep-Learning-based2026.06 | 66.9 | 65.4 | |
| MUSIQTraining dataset=KADID-10K, Method category=Discriminative Deep-Learning-based2026.06 | 65.6 | 56.3 | |
| BRISQUETraining dataset=KADID-10K, Method category=Handcrafted2026.06 | 38.9 | 38 | |
| NIQETraining dataset=KADID-10K, Method category=Handcrafted2026.06 | 36 | 37.6 |