Code problems

[slk1120]

Thank you for your excellent work！I encountered a few problems when replicating the code, and hope to get your answers, thanks!

1. In the step3, augmentation features need to be extracted, for example in datasets/cutmix_feature_extraction.py, SemiADNet code is not provided, is it a DRA or DevNet model?
2. In addition, the feature and featuer_scale extraction processes are not provided in the code. Therefore, I need to write a script to extract and save the original features？
3. In lines 60-68 of AHL-main/datasets/mvtecad.py, it does not seem to properly group data belonging to the same cluster in normal_data. In the internal loop, it always obtains data from the start position of normal_data, rather than the index in self.cluster_index[i], which affects the subsequent construction process of support set - query set pairs.
   @Diana1026

[changchunteng]

Hello, I have encountered each of your problems when reproducing them. Have you solved them？