Comprehensive Image Fusion Evaluation on MSRS (test)
1.079VIFHybrid Fusion
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Hybrid FusionTraining Data=MSRS Full Image, Training Epochs=100, Inference Mode=Full2026.02 | 1.079 | 0.721 | 57 | — | 17.264 | |
| DTPFTraining Data=MSRS Patch (128×128), Training Epochs=500, Inference Mode=Slide + LLM2026.02 | 1.06 | 0.732 | 13.1 | — | 40.3 | |
| Text-IFTraining Data=4 Datasets Patch (96×96), Inference Mode=Full + LLM2026.02 | 1.046 | 0.676 | — | — | 63.8 | |
| Hybrid FusionTraining Data=MSRS Full Image, Training Epochs=10, Inference Mode=Full2026.02 | 1.017 | 0.691 | 5.7 | — | 17.264 | |
| Hybrid FusionTraining Data=MSRS Full Image, Training Epochs=2, Inference Mode=Full2026.02 | 0.996 | 0.678 | 1.2 | — | 17.264 | |
| SwinFusionTraining Data=MSRS Patch (128×128), Training Epochs=120, Inference Mode=Full2026.02 | 0.825 | 0.558 | 2 | — | 13.04 | |
| CDDFuseTraining Data=MSRS Patch (128×128), Training Epochs=120, Inference Mode=Full2026.02 | 0.819 | 0.548 | 1 | — | 2.4 | |
| SuperFusionTraining Data=MSRS Patch (256×256), Training Epochs=1200, Inference Mode=Full2026.02 | 0.813 | 0.557 | 10 | — | 11.23 | |
| TC-MOATraining Data=MSRS Patch (448×448), Inference Mode=Pad-6402026.02 | 0.811 | 0.565 | 5 | — | 7.24 | |
| DDFMTraining Data=ILSVRC (Pre-trained), Inference Mode=Zero-shot Full (100 step)2026.02 | 0.81 | 0.58 | — | — | 552.81 | |
| PSLPTTraining Data=MSRS Patch (128×128 rep.), Inference Mode=Full2026.02 | 0.753 | 0.553 | 1 | — | 3.06 | |
| IFCNNTraining Data=MSRS Patch (128×128), Training Epochs=120, Inference Mode=Full2026.02 | 0.579 | 0.479 | 1 | — | 0.08 | |
| U2FusionTraining Data=MSRS Patch (64×64), Training Epochs=120, Inference Mode=Full2026.02 | 0.512 | 0.391 | 1 | — | 0.63 |