yunbo426 / predrnn-pp Goto Github PK

Hello, thank you for the paper and releasing the code. I want to reproduce predawn-pp on the kth action dataset using the parameters which provided in e3d, but I can't get the results of the paper which report on the kth task.
Maybe I'm not running it the right way, could you give me some advice ,thanks.

I am trying to use your code for video prediction with large resolution, like 1281281, and I wonder how much RAM will it consume? Have you tried some tricks, like inserting pooling and unpooling layers to reduce the RAM consumption?
Also, I have found in your paper that there is another experiment on KTH data set. Do you have information about the configuration of hyperparameters, RAM consumption and training speed of that experiment?

could you please provide me with an example for this two lists

i wanted to run it on my own dataset but i am lost. my data is of 6 videos with 57 frame at each video and each frame is 100x100 and the channel is 1 : (6,57,100,100,1)

i need help please?

Dear wang,

Could you give me some suggetions for predrnn used in speech processing?

many thanks