ValueError: Error when checking input: expected input_12 to have 4 dimensions, but got array with shape (4, 1)

[sandhya9173]

I mnot able to understand where this problem in this line its showing error(d_loss_real = discriminator.train_on_batch(image_batch_hr, real_data_Y) but my images size(shape (96, 96, 3)), can you please where its problem

[sandhya9173]

Both check please help in this error I mnot understanding

[deepak112]

You still facing this issue?

[sirius0503]

@deepak112 Hi Deepak, I am facing a similar issue could you please look at this thread here, I think it is a similar issue, except @sandhya9173 is trying to train the model, whereas I am trying to evaluate your pretrained model issue

[mustaeenqazi]

@deepak112 Hi Deepak iam facing the same issue while training, using the simplified code, in Colab

[deepak112]

Hi,
Can you please tell me shape of x_train_lr, x_train_hr, x_test_lr, x_test_hr in your code.

[mustaeenqazi]

thanks for prompt reply, here are the shapes:
x_train_lr = (500,)
x_train_hr = (500,)
x_test_lr = (300,)
x_test_lr = (300,)

[deepak112]

Are all images are of same shape?

[mustaeenqazi]

[mustaeenqazi]

shall i use the function load_data_from_dirs_resize instead of load_data_from_dirs, to load the data?

[deepak112]

Ya images are not of same size. Either make one custom funtion to have same size images. And then save those images to some directory and then use load_data_from_dirs.