search

xrli-U / MuSc

This is an official PyTorch implementation for "MuSc : Zero-Shot Industrial Anomaly Classification and Segmentation with Mutual Scoring of the Unlabeled Images" (MuSc ICLR2024).
MIT License
207 stars 12 forks source link

A little question #4

Open NAZABANMAAFK opened 3 months ago

NAZABANMAAFK commented 3 months ago

Why is the test image scaled to a resolution size of 518*518, is there any reason or prior knowledge?

xrli-U commented 3 months ago

We adopted the same setting as in the APRIL-GAN method to keep the comparison fair. When image size is larger, each patch can contain the finer contents, which is conducive to segment the small and complex abnormal regions. If you want to know more about the results of different image sizes, please see the Results of different backbones in README.md.