search

cvlab-yonsei / MNAD

An official implementation of "Learning Memory-guided Normality for Anomaly Detection" (CVPR 2020) in PyTorch.
333 stars 82 forks source link

How to decide number of memory items "m" ? #18

Closed UsamaI000 closed 3 years ago

UsamaI000 commented 3 years ago

Hi, great work. There is something I would like to know. How do you decide the number of memory items? Like in your paper you used 10 memory items How do you decide that number ? Is there a rationale behind it ? I'll be very grateful to get an answer. Thank you

hyunjp commented 3 years ago

We implemented ablation study according to the number of the memory items. We found that the experiment using 10 memory items gives the best performance.