search

michaeltrs / DeepSatModels

Deep learning models for remote sensing applications
Apache License 2.0
151 stars 27 forks source link

How to inference for PASTIS24 #7

Open vvuonghn opened 1 year ago

vvuonghn commented 1 year ago

Hi @michaeltrs

Thank you for your release source code.

I have trained the model for fold 1 PASTIS24. The process was completed, but I don't know how to infer the model and merge it into original size HxW (128x128)

Also, There are a fews syntax bug, such as for j in unfolded_images.shape[0]: link

rongtongxueya commented 1 year ago

I'm currently working on time series, but a question has been bugging me: how do you visualize the extrapolation of a CHW image using the saved BCTHW's model, and do you visualize the entire time series as a result of the extrapolation? Or do you replicate the image in the time dimension for this moment in time?

PeterKKan commented 7 months ago

should be: for j in range(unfolded_images.shape[0]):

IwantNewway commented 6 months ago

I'm currently working on time series, but a question has been bugging me: how do you visualize the extrapolation of a CHW image using the saved BCTHW's model, and do you visualize the entire time series as a result of the extrapolation? Or do you replicate the image in the time dimension for this moment in time?

Do u fix it? i meet same question? thanks!