def __call__(self, y_true, y_pred):
# There are additional parameters for this function
# Note: some of the 'modes' for edge behavior do not yet have a
# gradient definition in the Theano tree
# and cannot be used for learning
kernel = [self.kernel_size, self.kernel_size]
y_true = K.reshape(y_true, [-1] + list(self.__int_shape(y_pred)[1:]))
y_pred = K.reshape(y_pred, [-1] + list(self.__int_shape(y_pred)[1:]))
patches_pred = KC.extract_image_patches(y_pred, kernel, kernel,
'valid', self.dim_ordering)
patches_true = KC.extract_image_patches(y_true, kernel, kernel,
'valid', self.dim_ordering)
# Reshape to get the var in the cells
bs, w, h, c1, c2, c3 = self.__int_shape(patches_pred)
patches_pred = K.reshape(patches_pred, [-1, w, h, c1 * c2 * c3])
patches_true = K.reshape(patches_true, [-1, w, h, c1 * c2 * c3])
# Get mean
u_true = K.mean(patches_true, axis=-1)
u_pred = K.mean(patches_pred, axis=-1)
# Get variance
var_true = K.var(patches_true, axis=-1)
var_pred = K.var(patches_pred, axis=-1)
# Get std dev
covar_true_pred = K.mean(patches_true * patches_pred,
axis=-1) - u_true * u_pred
ssim = (2 * u_true * u_pred + self.c1) * (2 * covar_true_pred +
self.c2)
denom = ((K.square(u_true) + K.square(u_pred) + self.c1) *
(var_pred + var_true + self.c2))
ssim /= denom # no need for clipping, c1 and c2 make the denom non-zero
return K.mean((1.0 - ssim) / 2.0)
def DSSIM_base(y_true, y_pred, kernel_size):
kernel = [kernel_size, kernel_size]
y_true = K.reshape(y_true, [-1] + list(K.int_shape(y_pred)[1:]))
y_pred = K.reshape(y_pred, [-1] + list(K.int_shape(y_pred)[1:]))
patches_pred = KC.extract_image_patches(y_pred, kernel, kernel, 'valid',
K.image_data_format())
patches_true = KC.extract_image_patches(y_true, kernel, kernel, 'valid',
K.image_data_format())
# Reshape to get the var in the cells
bs, w, h, c1, c2, c3 = K.int_shape(patches_pred)
patches_pred = K.reshape(patches_pred, [-1, w, h, c1 * c2 * c3])
patches_true = K.reshape(patches_true, [-1, w, h, c1 * c2 * c3])
# Get mean
u_true = K.mean(patches_true, axis=-1)
u_pred = K.mean(patches_pred, axis=-1)
# Get variance
var_true = K.var(patches_true, axis=-1)
var_pred = K.var(patches_pred, axis=-1)
# Get std dev
covar_true_pred = K.mean(patches_true * patches_pred,
axis=-1) - u_true * u_pred
ssim = (2 * u_true * u_pred + 0.01**2) * (2 * covar_true_pred + 0.03**2)
denom = ((K.square(u_true) + K.square(u_pred) + 0.01**2) *
(var_pred + var_true + 0.03**2))
ssim /= denom # no need for clipping, c1 and c2 make the denom non-zero
print('----------------------', K.mean((1.0 - ssim) / 2.0))
return K.mean((1.0 - ssim) / 2.0)
def SSIM(y_true, y_pred):
# There are additional parameters for this function
# Note: some of the 'modes' for edge behavior do not yet have a gradient definition in the Theano tree
# and cannot be used for learning
kernel_size = 3
k1 = 0.01
k2 = 0.03
max_value = 1.0
c1 = (k1 * max_value)**2
c2 = (k2 * max_value)**2
dim_ordering = K.image_data_format()
#backend = KC.backend()
kernel = [kernel_size, kernel_size]
y_true = KC.reshape(y_true, [-1] + list(int_shape(y_pred)[1:]))
y_pred = KC.reshape(y_pred, [-1] + list(int_shape(y_pred)[1:]))
patches_pred = KC.extract_image_patches(y_pred, kernel, kernel, 'valid',
dim_ordering)
patches_true = KC.extract_image_patches(y_true, kernel, kernel, 'valid',
dim_ordering)
# Reshape to get the var in the cells
bs, w, h, c1, c2, c3 = int_shape(patches_pred)
patches_pred = KC.reshape(patches_pred, [-1, w, h, c1 * c2 * c3])
patches_true = KC.reshape(patches_true, [-1, w, h, c1 * c2 * c3])
# Get mean
u_true = KC.mean(patches_true, axis=-1)
u_pred = KC.mean(patches_pred, axis=-1)
# Get variance
var_true = K.var(patches_true, axis=-1)
var_pred = K.var(patches_pred, axis=-1)
# Get std dev
covar_true_pred = K.mean(patches_true * patches_pred,
axis=-1) - u_true * u_pred
ssim = (2 * u_true * u_pred + c1) * (2 * covar_true_pred + c2)
denom = (K.square(u_true) + K.square(u_pred) + c1) * (var_pred + var_true +
c2)
ssim /= denom # no need for clipping, c1 and c2 make the denom non-zero
return ssim #K.mean((1.0 - ssim) / 2.0)
def ssim_loss(y_true, y_pred):
kernel = [3, 3]
k1 = 0.01
k2 = 0.03
kernel_size = 3
max_value = 1.0
cc1 = (k1 * max_value)**2
cc2 = (k2 * max_value)**2
y_true = K.reshape(y_true, [-1] + list(K.int_shape(y_pred)[1:]))
y_pred = K.reshape(y_pred, [-1] + list(K.int_shape(y_pred)[1:]))
patches_pred = KC.extract_image_patches(y_pred, kernel, kernel, 'valid',
K.image_data_format())
patches_true = KC.extract_image_patches(y_true, kernel, kernel, 'valid',
K.image_data_format())
bs, w, h, c1, c2, c3 = K.int_shape(patches_pred)
patches_pred = K.reshape(patches_pred, [-1, w, h, c1 * c2 * c3])
patches_true = K.reshape(patches_true, [-1, w, h, c1 * c2 * c3])
# Get mean
u_true = K.mean(patches_true, axis=-1)
u_pred = K.mean(patches_pred, axis=-1)
# Get variance
var_true = K.var(patches_true, axis=-1)
var_pred = K.var(patches_pred, axis=-1)
# Get covariance
covar_true_pred = K.mean(patches_true * patches_pred,
axis=-1) - u_true * u_pred
ssim = (2 * u_true * u_pred + cc1) * (2 * covar_true_pred + cc2)
denom = (K.square(u_true) + K.square(u_pred) + cc1) * \
(var_pred + var_true + cc2)
ssim /= denom
return K.mean((1.0 - ssim) / 2.0)
