def main(args):
r"""Performs the main training loop
"""
# Load dataset
print('> Loading dataset ...')
dataset_train = Dataset(train=True, gray_mode=args.gray, shuffle=True)
dataset_val = Dataset(train=False, gray_mode=args.gray, shuffle=False)
loader_train = DataLoader(dataset=dataset_train, num_workers=6, \
batch_size=args.batch_size, shuffle=True)
print("\t# of training samples: %d\n" % int(len(dataset_train)))
# Init loggers
if not os.path.exists(args.log_dir):
os.makedirs(args.log_dir)
writer = SummaryWriter(args.log_dir)
logger = init_logger(args)
# Create model
if not args.gray:
in_ch = 3
else:
in_ch = 1
net = FFDNet(num_input_channels=in_ch)
# Initialize model with He init
net.apply(weights_init_kaiming)
# Define loss
criterion = nn.MSELoss(size_average=False)
# Move to GPU
device_ids = [0]
model = nn.DataParallel(net, device_ids=device_ids).cuda()
criterion.cuda()
# Optimizer
optimizer = optim.Adam(model.parameters(), lr=args.lr)
# Resume training or start anew
if args.resume_training:
resumef = os.path.join(args.log_dir, 'ckpt.pth')
if os.path.isfile(resumef):
checkpoint = torch.load(resumef)
print("> Resuming previous training")
model.load_state_dict(checkpoint['state_dict'])
optimizer.load_state_dict(checkpoint['optimizer'])
new_epoch = args.epochs
new_milestone = args.milestone
current_lr = args.lr
args = checkpoint['args']
training_params = checkpoint['training_params']
start_epoch = training_params['start_epoch']
args.epochs = new_epoch
args.milestone = new_milestone
args.lr = current_lr
print("=> loaded checkpoint '{}' (epoch {})"\
.format(resumef, start_epoch))
print("=> loaded parameters :")
print("==> checkpoint['optimizer']['param_groups']")
print("\t{}".format(checkpoint['optimizer']['param_groups']))
print("==> checkpoint['training_params']")
for k in checkpoint['training_params']:
print("\t{}, {}".format(k, checkpoint['training_params'][k]))
argpri = vars(checkpoint['args'])
print("==> checkpoint['args']")
for k in argpri:
print("\t{}, {}".format(k, argpri[k]))
args.resume_training = False
else:
raise Exception("Couldn't resume training with checkpoint {}".\
format(resumef))
else:
start_epoch = 0
training_params = {}
training_params['step'] = 0
training_params['current_lr'] = 0
training_params['no_orthog'] = args.no_orthog
# Training
for epoch in range(start_epoch, args.epochs):
# Learning rate value scheduling according to args.milestone
if epoch > args.milestone[1]:
current_lr = args.lr / 1000.
training_params['no_orthog'] = True
elif epoch > args.milestone[0]:
current_lr = args.lr / 10.
else:
current_lr = args.lr
# set learning rate in optimizer
for param_group in optimizer.param_groups:
param_group["lr"] = current_lr
print('learning rate %f' % current_lr)
# train
for i, data in enumerate(loader_train, 0):
# Pre-training step
model.train()
model.zero_grad()
optimizer.zero_grad()
# inputs: noise and noisy image
img_train = data
noise = torch.zeros(img_train.size())
stdn = np.random.uniform(args.noiseIntL[0], args.noiseIntL[1], \
size=noise.size()[0])
for nx in range(noise.size()[0]):
sizen = noise[0, :, :, :].size()
noise[nx, :, :, :] = torch.FloatTensor(sizen).\
normal_(mean=0, std=stdn[nx])
imgn_train = img_train + noise
# Create input Variables
img_train = Variable(img_train.cuda(), volatile=True)
imgn_train = Variable(imgn_train.cuda(), volatile=True)
noise = Variable(noise.cuda())
stdn_var = Variable(torch.cuda.FloatTensor(stdn), volatile=True)
# Evaluate model and optimize it
out_train = model(imgn_train, stdn_var)
loss = criterion(out_train, noise) / (imgn_train.size()[0] * 2)
loss.backward()
optimizer.step()
# Results
model.eval()
out_train = torch.clamp(imgn_train - model(imgn_train, stdn_var),
0., 1.)
psnr_train = batch_psnr(out_train, img_train, 1.)
# PyTorch v0.4.0: loss.data[0] --> loss.item()
if training_params['step'] % args.save_every == 0:
# Apply regularization by orthogonalizing filters
if not training_params['no_orthog']:
model.apply(svd_orthogonalization)
# Log the scalar values
writer.add_scalar('loss', loss.data[0],
training_params['step'])
writer.add_scalar('PSNR on training data', psnr_train, \
training_params['step'])
print("[epoch %d][%d/%d] loss: %.4f PSNR_train: %.4f" %\
(epoch+1, i+1, len(loader_train), loss.data[0], psnr_train))
training_params['step'] += 1
# The end of each epoch
model.eval()
# Validation
psnr_val = 0
for valimg in dataset_val:
img_val = torch.unsqueeze(valimg, 0)
noise = torch.FloatTensor(img_val.size()).\
normal_(mean=0, std=args.val_noiseL)
imgn_val = img_val + noise
img_val, imgn_val = Variable(img_val.cuda()), Variable(
imgn_val.cuda())
sigma_noise = Variable(torch.cuda.FloatTensor([args.val_noiseL]))
out_val = torch.clamp(imgn_val - model(imgn_val, sigma_noise), 0.,
1.)
psnr_val += batch_psnr(out_val, img_val, 1.)
psnr_val /= len(dataset_val)
print("\n[epoch %d] PSNR_val: %.4f" % (epoch + 1, psnr_val))
writer.add_scalar('PSNR on validation data', psnr_val, epoch)
writer.add_scalar('Learning rate', current_lr, epoch)
# Log val images
try:
if epoch == 0:
# Log graph of the model
writer.add_graph(
model,
(imgn_val, sigma_noise),
)
# Log validation images
for idx in range(2):
imclean = utils.make_grid(img_val.data[idx].clamp(0., 1.), \
nrow=2, normalize=False, scale_each=False)
imnsy = utils.make_grid(imgn_val.data[idx].clamp(0., 1.), \
nrow=2, normalize=False, scale_each=False)
writer.add_image('Clean validation image {}'.format(idx),
imclean, epoch)
writer.add_image('Noisy validation image {}'.format(idx),
imnsy, epoch)
for idx in range(2):
imrecons = utils.make_grid(out_val.data[idx].clamp(0., 1.), \
nrow=2, normalize=False, scale_each=False)
writer.add_image('Reconstructed validation image {}'.format(idx), \
imrecons, epoch)
# Log training images
imclean = utils.make_grid(img_train.data, nrow=8, normalize=True, \
scale_each=True)
writer.add_image('Training patches', imclean, epoch)
except Exception as e:
logger.error("Couldn't log results: {}".format(e))
# save model and checkpoint
training_params['start_epoch'] = epoch + 1
torch.save(model.state_dict(), os.path.join(args.log_dir, 'net.pth'))
save_dict = { \
'state_dict': model.state_dict(), \
'optimizer' : optimizer.state_dict(), \
'training_params': training_params, \
'args': args\
}
torch.save(save_dict, os.path.join(args.log_dir, 'ckpt.pth'))
if epoch % args.save_every_epochs == 0:
torch.save(save_dict, os.path.join(args.log_dir, \
'ckpt_e{}.pth'.format(epoch+1)))
del save_dict
def test_ffdnet(**args):
r"""Denoises an input image with FFDNet
"""
# Init logger
logger = init_logger_ipol()
# Check if input exists and if it is RGB
try:
rgb_den = is_rgb(args['input'])
except:
raise Exception('Could not open the input image')
# Open image as a CxHxW torch.Tensor
if rgb_den:
in_ch = 3
model_fn = 'models/net_rgb.pth'
imorig = cv2.imread(args['input'])
# from HxWxC to CxHxW, RGB image
imorig = (cv2.cvtColor(imorig, cv2.COLOR_BGR2RGB)).transpose(2, 0, 1)
else:
# from HxWxC to CxHxW grayscale image (C=1)
in_ch = 1
model_fn = 'logs/net.pth'
imorig = cv2.imread(args['input'], cv2.IMREAD_GRAYSCALE)
imorig_copy = imorig.copy()
imorig = np.expand_dims(imorig, 0)
imorig = np.expand_dims(imorig, 0)
# Handle odd sizes
expanded_h = False
expanded_w = False
sh_im = imorig.shape
if sh_im[2] % 2 == 1:
expanded_h = True
imorig = np.concatenate((imorig, \
imorig[:, :, -1, :][:, :, np.newaxis, :]), axis=2)
if sh_im[3] % 2 == 1:
expanded_w = True
imorig = np.concatenate((imorig, \
imorig[:, :, :, -1][:, :, :, np.newaxis]), axis=3)
imorig = normalize(imorig)
imorig = torch.Tensor(imorig)
# Absolute path to model file
model_fn = os.path.join(os.path.abspath(os.path.dirname(__file__)), \
model_fn)
# Create model
print('Loading model ...\n')
net = FFDNet(num_input_channels=in_ch)
# Load saved weights
if args['cuda']:
state_dict = torch.load(model_fn)
device_ids = [0]
model = nn.DataParallel(net, device_ids=device_ids).cuda()
else:
state_dict = torch.load(model_fn, map_location='cpu')
# CPU mode: remove the DataParallel wrapper
state_dict = remove_dataparallel_wrapper(state_dict)
model = net
model.load_state_dict(state_dict)
# Sets the model in evaluation mode (e.g. it removes BN)
model.eval()
# Sets data type according to CPU or GPU modes
if args['cuda']:
dtype = torch.cuda.FloatTensor
else:
dtype = torch.FloatTensor
# Add noise
if args['add_noise']:
noise = torch.FloatTensor(imorig.size()).\
normal_(mean=0, std=args['noise_sigma'])
imnoisy = imorig + noise
else:
imnoisy = imorig.clone()
# Test mode
with torch.no_grad(): # PyTorch v0.4.0
imorig, imnoisy = Variable(imorig.type(dtype)), \
Variable(imnoisy.type(dtype))
nsigma = Variable(torch.FloatTensor([args['noise_sigma']]).type(dtype))
# Measure runtime
start_t = time.time()
# Estimate noise and subtract it to the input image
im_noise_estim = model(imnoisy, nsigma)
outim = torch.clamp(imnoisy - im_noise_estim, 0., 1.)
stop_t = time.time()
if expanded_h:
imorig = imorig[:, :, :-1, :]
outim = outim[:, :, :-1, :]
imnoisy = imnoisy[:, :, :-1, :]
if expanded_w:
imorig = imorig[:, :, :, :-1]
outim = outim[:, :, :, :-1]
imnoisy = imnoisy[:, :, :, :-1]
# Compute PSNR and log it
if rgb_den:
print("### RGB denoising ###")
else:
print("### Grayscale denoising ###")
if args['add_noise']:
psnr = batch_psnr(outim, imorig, 1.)
psnr_noisy = batch_psnr(imnoisy, imorig, 1.)
print("----------PSNR noisy {0:0.2f}dB".format(psnr_noisy))
print("----------PSNR denoised {0:0.2f}dB".format(psnr))
else:
logger.info("\tNo noise was added, cannot compute PSNR")
print("----------Runtime {0:0.4f}s".format(stop_t - start_t))
# Compute difference
diffout = 2 * (outim - imorig) + .5
diffnoise = 2 * (imnoisy - imorig) + .5
# Save images
if not args['dont_save_results']:
noisyimg = variable_to_cv2_image(imnoisy)
outimg = variable_to_cv2_image(outim)
cv2.imwrite(
"bfffd/noisy-" + str(int(args['noise_sigma'] * 255)) + '-' +
args['input'], noisyimg)
cv2.imwrite(
"bfffd/ffdnet-" + str(int(args['noise_sigma'] * 255)) + '-' +
args['input'], outimg)
if args['add_noise']:
cv2.imwrite("noisy_diff.png", variable_to_cv2_image(diffnoise))
cv2.imwrite("ffdnet_diff.png", variable_to_cv2_image(diffout))
(score, diff) = compare_ssim(noisyimg, imorig_copy, full=True)
(score2, diff) = compare_ssim(outimg, imorig_copy, full=True)
print("----------Noisy ssim: {0:0.4f}".format(score))
print("----------Denoisy ssim: {0:0.4f}".format(score2))
# cuda = True
cuda = False
# Only Grayscale model is provided here, the RGB model is even better.
rgb_den = False
B = 2048.
# B = 256.
########
in_ch = 1
# This model was trained on the Waterloo Exploration Database, sigma: [0,20], 50 epochs. see Fig.2 of the paper
# It cost about 12 hours in a GPU server (Nvidia RTX 2080Ti with 11GB graphic memory)
model_fn = 'mat/net-gray-v2.pth'
# Absolute path to model file
model_fn = os.path.join(os.path.abspath(os.path.dirname(__file__)), model_fn)
# Create model
print('Loading model ...\n')
net = FFDNet(num_input_channels=in_ch)
# Load saved weights
if cuda:
state_dict = torch.load(model_fn)
device_ids = [0]
model = nn.DataParallel(net, device_ids=device_ids).cuda()
# Sets data type according to CPU or GPU modes
dtype = torch.cuda.FloatTensor
else:
state_dict = torch.load(model_fn, map_location='cpu')
# CPU mode: remove the DataParallel wrapper
state_dict = remove_dataparallel_wrapper(state_dict)
model = net
dtype = torch.FloatTensor
model.load_state_dict(state_dict)
def test_ffdnet(**args):
r"""Denoises an input image with FFDNet
"""
# Init logger
logger = init_logger_ipol()
# Check if input exists and if it is RGB
try:
rgb_den = is_rgb(args['input'])
except:
raise Exception('Could not open the input image')
# Measure runtime
start_t = time.time()
# Open image as a CxHxW torch.Tensor
if rgb_den:
in_ch = 3
model_fn = 'net_rgb.pth'
imorig = Image.open(args['input'])
imorig = np.array(imorig, dtype=np.float32).transpose(2, 0, 1)
else:
# from HxWxC to CxHxW grayscale image (C=1)
in_ch = 1
model_fn = 'models/net_gray.pth'
# imorig = cv2.imread(args['input'], cv2.IMREAD_GRAYSCALE)
imorig = np.expand_dims(imorig, 0)
imorig = np.expand_dims(imorig, 0)
# Handle odd sizes
expanded_h = False
expanded_w = False
sh_im = imorig.shape
if sh_im[2] % 2 == 1:
expanded_h = True
imorig = np.concatenate((imorig, \
imorig[:, :, -1, :][:, :, np.newaxis, :]), axis=2)
if sh_im[3] % 2 == 1:
expanded_w = True
imorig = np.concatenate((imorig, \
imorig[:, :, :, -1][:, :, :, np.newaxis]), axis=3)
imorig = normalize(imorig)
imorig = torch.Tensor(imorig)
# Absolute path to model file
model_fn = os.path.join(os.path.abspath(os.path.dirname(__file__)), \
model_fn)
# Create model
print('Loading model ...\n')
net = FFDNet(num_input_channels=in_ch)
# Load saved weights
if args['cuda']:
state_dict = torch.load(model_fn)
#device_ids = [0,1,2,3]
#model = nn.DataParallel(net, device_ids=device_ids).cuda()
#state_dict = remove_dataparallel_wrapper(state_dict)
model = net
else:
state_dict = torch.load(model_fn, map_location='cpu')
# CPU mode: remove the DataParallel wrapper
state_dict = remove_dataparallel_wrapper(state_dict)
model = net
model.load_state_dict(state_dict)
# Sets the model in evaluation mode (e.g. it removes BN)
model.eval()
# Sets data type according to CPU or GPU modes
if args['cuda']:
dtype = torch.cuda.FloatTensor
else:
dtype = torch.FloatTensor
# Test mode
with torch.no_grad(): # PyTorch v0.4.0
imorig = Variable(imorig.type(dtype))
nsigma = Variable(torch.FloatTensor([args['noise_sigma']]).type(dtype))
# # Measure runtime
# start_t = time.time()
# Estimate noise and subtract it to the input image
im_noise_estim = model(imorig, nsigma)
stop_t = time.time()
# log time
if rgb_den:
print("### RGB denoising ###")
else:
print("### Grayscale denoising ###")
print("\tRuntime {0:0.4f}s".format(stop_t - start_t))
# Save noises
noise = variable_to_numpy(imorig.to(3) - im_noise_estim).transpose(1, 2, 0)
filename = args['input'].split('/')[-1].split('.')[0]
if args['save_path']:
sio.savemat(
'./output_noise/' + args['save_path'] + '/' + filename + '.mat',
{'Noisex': noise})
else:
sio.savemat('./output_noise/' + filename + '.mat', {'Noisex': noise})
os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
os.environ["CUDA_VISIBLE_DEVICES"] = "0"
r"""Denoises an input image with FFDNet
"""
# Init logger
logger = init_logger_ipol()
# Check if input exists and if it is RGB
# Absolute path to model file
model_fn = os.path.join(os.path.abspath(os.path.dirname(__file__)), \
'models/net_rgb.pth')
# Create model
print('Loading model ...\n')
net = FFDNet(num_input_channels=3, test_mode=True)
model_fn = os.path.join(os.path.abspath(os.path.dirname(__file__)), \
model_fn)
# Load saved weights
print(model_fn)
state_dict = torch.load(model_fn)
device_ids = [0]
model = nn.DataParallel(net, device_ids=device_ids).cuda()
model.load_state_dict(state_dict)
# Sets the model in evaluation mode (e.g. it removes BN)
model.eval()
# Sets data type according to CPU or GPU modes
dtype = torch.cuda.FloatTensor