Open louis-xin opened 1 year ago
Have you solve the problem? I am also working on using A1,A2,A3 to predict B
Have you solve the problem? I am also working on using A1,A2,A3 to predict B
No not yet, do u have any idea for it?
Any update @workingpotato !?
Have you solve the problem? I am also working on using A1,A2,A3 to predict B
No not yet, do u have any idea for it?
I am also trying this but has same issue. I have stacked my input to 6 channel NPZ and changed the dataset loader accordingly. My output channels would be just 1. Has anyone found a solution for this ?
Have you solve the problem? I am also working on using multi-input but have no idea to solve it.
Hello I want to perform an A1,A2 --> B prediction using the cycle GAN model, as you mentioned in #498, I have to make a custom dataset to read the A1 and A2 image separately and concatenate it into (x,x,6) array A and pass it to the generator to generate an image B (x,x,3).
Problem is that the transform doesnt accept 6 channel np.array and also in the cycle_gan_model.py line 86, it only works when the input and output images have the same number of channels.
Could you give me some insights to solve this problem?
Thanks.
This is my custom dataset
`import os from data.base_dataset import BaseDataset, get_transform, get_params from data.image_folder import make_dataset from PIL import Image import random import numpy as np
from data.image_folder import make_dataset
from PIL import Image
class BucklingDataset(BaseDataset): """A template dataset class for you to implement custom datasets."""
@staticmethod