search

feiyuhuahuo / Anomaly_Prediction

Pytorch implementation of anomaly prediction.
78 stars 22 forks source link

About EER #6

Closed Tiacy closed 4 years ago

Tiacy commented 4 years ago

Hello，Could you tell me how to compute Equal Error Rate? By the way,do you have any idea to improve your prediction model? Have you tried to add a reconstruction model after the prediction model?

feiyuhuahuo commented 4 years ago

I don't know Equal Error Rate. You can set a threshold to distinguish the normal frame and the abnormal frame. To improve the model, a new and better GAN maybe is helpful.

Tiacy commented 4 years ago

OK.Thanks for your reply!

Tiacy commented 4 years ago

Another question , when training UCSD ped1/ped2 datasets,should we change "input_channels=12" into "input_channels=4", as the datasets are all grayscales and every image has one channel.

feiyuhuahuo commented 4 years ago

No need to change input_channels. Grey images will be treated as 3 channels. OpenCV will deal with it automatically.

Tiacy commented 4 years ago

Get it！Thanks for your help!