question about test result

YangSun22 / TC-MoA

Task-Customized Mixture of Adapters for General Image Fusion (CVPR 2024)

73 stars 8 forks source link

question about test result #15

Open sove45 opened 5 months ago

sove45 commented 5 months ago

When using the "Evaluation for single image" script to test the VIF fusion metrics on LLVIP, I unexpectedly found that the test results were significantly higher than those reported in the paper. The paper states an SD test value of 9.805, while my test results showed a value of 10.606

PLAlyj commented 2 months ago

您好，请问哪里有"Evaluation for single image"啊，没看到