Isn't there a mistake in the loss computation here?

WaqasSultani / AnomalyDetectionCVPR2018

502 stars 182 forks source link

Isn't there a mistake in the loss computation here? #52

Open mourad1081 opened 5 years ago

mourad1081 commented 5 years ago

https://github.com/WaqasSultani/AnomalyDetectionCVPR2018/blob/cadb6a1a98f8e8db41a91ca00fce5e1ab2ad8c8f/TrainingAnomalyDetector_public.py#L267

Isn't it supposed to be sub_z = T.maximum(1 - Sub_Abn[ii] + Sub_Nor[ii], 0)?

venkataganti commented 4 months ago

@mourad1081 Could you able to figure out the issue in the above line of code in loss function? Even I am stuck at the same point. If you could able to figure out the fix, can you please share the changes in the code. I appreciate it..