Open leolle520 opened 1 month ago
Our method uses a pre-trained image encoder of CLIP as the feature extractor, requiring no additional training and fine-tuning. In other words, most current feature extractors can be used in our method. All the feature extractors used in the script can be downloaded from the following link, CLIP: https://github.com/mlfoundations/open_clip DINO: https://github.com/facebookresearch/dino DINO_v2: https://github.com/facebookresearch/dinov2
If you want to test on your custom dataset, you can write an interface with the same output directly from def getitem(self, idx) in ./datasets/mvtec.py
How to apply this method on my custom dataset and do the prediction please? I don't see any training process, it's amazing.
Thank you.