def forward(self, y, true, it, writer=None):
n = self.sizes[0]
y_current = nn.functional.interpolate(y, self.proj_shapes[0], mode='area')
current = self.init_op(y_current)
curs = []
grads = []
fbpys = []
for i in range(len(self.sizes)):
n = self.sizes[i]
proj_shape = self.proj_shapes[i]
if i < len(self.sizes)-1:
y_current = nn.functional.interpolate(y, proj_shape, mode='area')
else:
y_current = y
if i < 10:
current = nn.functional.interpolate(current, (n, n), mode='bilinear')
iteration = getattr(self, 'iteration_{}'.format(i))
current = iteration(current,curs,grads,fbpys, y_current,i)
curs.append(current)
else:
current = nn.functional.interpolate(current, (n, n), mode='bilinear')
iteration = getattr(self, 'iteration_{}'.format(i))
current = iteration(current,curs,grads,fbpys, y_current,i)
if writer:
util.summary_image(writer, 'iteration_{}'.format(i), current, it)
return current, self.loss(current, true)
def summaries(writer, result, fbp, true, loss, it, do_print=False):
"""Save and print training and validation data to tensorboard"""
residual = result - true
squared_error = residual**2
mse = torch.mean(squared_error)
maxval = torch.max(true) - torch.min(true)
psnr = 20 * torch.log10(maxval) - 10 * torch.log10(mse)
# ratio of mse to fbp mse
relative = torch.mean((result - true)**2) / torch.mean((fbp - true)**2)
ssi = ssim(result, true)
ssi_fbp = ssim(fbp, true)
# relative ssim
relative_ssim = ssi / ssi_fbp
if do_print:
print(it, mse.item(), psnr.item(), relative.item(), ssi.item(),
relative_ssim.item())
writer.add_scalar('loss', loss, it)
writer.add_scalar('psnr', psnr, it)
writer.add_scalar('relative', relative, it)
writer.add_scalar('ssim', ssi, it)
writer.add_scalar('relative ssim', relative_ssim, it)
util.summary_image(writer, 'result', result, it)
util.summary_image(writer, 'true', true, it)
def forward(self, y, true, it, writer=None):
current = self.init_op(y)
for i in range(self.niter):
iteration = getattr(self, 'iteration_{}'.format(i))
current = iteration(current, y)
if writer:
util.summary_image(writer, 'iteration_{}'.format(i), current, it)
return current, self.loss(current, true)
def summaries(writer, result, fbp, true, loss, it, do_print=False):
residual = result - true
squared_error = residual ** 2
mse = torch.mean(squared_error)
maxval = torch.max(true) - torch.min(true)
psnr = 20 * torch.log10(maxval) - 10 * torch.log10(mse)
relative = torch.mean((result - true) ** 2) / torch.mean((fbp - true) ** 2)
if do_print:
print(it, mse.item(), psnr.item(), relative.item())
writer.add_scalar('loss', loss, it)
writer.add_scalar('psnr', psnr, it)
writer.add_scalar('relative', relative, it)
util.summary_image(writer, 'result', result, it)
util.summary_image(writer, 'true', true, it)
