Do you have a link to download the data set? Can you explain the training steps?

[fangliangs]

No description provided.

[zhaipro]

图片是爬取来的，具体来源基本都忘了。mask是调用美图获取的。
https://drive.google.com/drive/folders/1SFZgEaw0x3Z8F7_-SgVz5zs2-iTaTMPa?usp=sharing

[fangliangs]

ERR when checking input: expected input_1 to have 4 dimensions, but got array with shape (1, 1)

def data_generator(path, batch_size=2):
'''data generator for fit_generator'''
n = 3000
i = 0
idxs = np.arange(n)
while True:
x, y = [], []
for b in range(batch_size):
if i % n == 0:
np.random.shuffle(idxs)
fn = f'{path}/x/{idxs[i]}.jpg'
im = cv2.imread(fn)
x.append(im)
fn = f'{path}/y/{idxs[i]}.jpg'
im = cv2.imread(fn, 0)
y.append(im)
i = (i + 1) % n
x = np.array(x, dtype='float32') / 255
y = np.array(y, dtype='float32') / 255
y.shape = y.shape + (1,)
yield x, y