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

display abnormal region #27

Open pankSM opened 3 years ago

pankSM commented 3 years ago

Thanks for the perfect code. I want to display the prediction errot,how can I display it? thanks for any suggestion!

Pokeysx commented 3 years ago

Can you share some of the problems you have encountered while running code？

pankSM commented 3 years ago

first, the gpu memory is increasing during test; second, in paper, For visualizing the anomalies, we compute pixel-wise abnormality scores similar to (16). this part should be a const value, how to measture each pixel value; third, I want to display mask and to show it in corresponding image.but I don't find this code.can you share the code. my email is:1443922774@qq.com Thanks.

xiximin commented 3 years ago

Can you share the shanghaitech dataset. I always fail to download.

pankSM commented 3 years ago

sorry, I don't have the dataset

Be-a-strong-man commented 2 years ago

first, the gpu memory is increasing during test; second, in paper, For visualizing the anomalies, we compute pixel-wise abnormality scores similar to (16). this part should be a const value, how to measture each pixel value; third, I want to display mask and to show it in corresponding image.but I don't find this code.can you share the code. my email is:1443922774@qq.com Thanks.

hi, do you receive the code that can visualize the anomalies, if you receive it. would you please give me this code, thanks a lot. my email is: wjh_0610@163.com

yachuchi commented 5 months ago

first, the gpu memory is increasing during test; second, in paper, For visualizing the anomalies, we compute pixel-wise abnormality scores similar to (16). this part should be a const value, how to measture each pixel value; third, I want to display mask and to show it in corresponding image.but I don't find this code.can you share the code. my email is:1443922774@qq.com Thanks.

hi, do you receive the code that can visualize the anomalies, if you receive it. would you please give me this code, thanks a lot. my email is: wjh_0610@163.com

Hi, I want the code for the visualization part. If you receive it, would you please give me this code? My mail: julie.gee1211@gmail.com