search

plutoyuxie / AutoEncoder-SSIM-for-unsupervised-anomaly-detection-

Improving Unsupervised Defect Segmentation by Applying Structural Similarity to Autoencoders
121 stars 25 forks source link

License question #8

Open bousi618 opened 3 years ago

bousi618 commented 3 years ago

Please tell me the license of this repository

XuBlack commented 3 years ago

how do you choose the threshhold?

plutoyuxie commented 3 years ago

how do you choose the threshhold?

Hi, the threshold is estimated based on reconstruction errors of valid positive samples. Please check the get_threshold function in test.py for details.