def main():
'''
main function, start point
'''
# 引数関連
parser = argparse.ArgumentParser()
parser.add_argument('--batchsize',
'-b',
type=int,
default=128,
help='Number of images in each mini-batch')
parser.add_argument('--gpu',
'-g',
type=int,
default=0,
help='Number of images in each mini-batch')
parser.add_argument('--model', '-m', type=str, help=('Using Model'))
args = parser.parse_args()
# parameter出力
# 保存ディレクトリ
# save didrectory
# prepare model
# 個々でネットワークを帰る
model = N.GenEvaluator(N.AEFINetConcat(ch=4, f_size=5))
if args.gpu >= 0:
chainer.cuda.get_device_from_id(args.gpu).use()
model.to_gpu()
model_path = path.join(ROOT_PATH, 'models', args.model)
chainer.serializers.load_npz(model_path, model)
movie_img = 'acrobat_s01'
img_list = sorted(os.listdir(path.join(FULL_PATH, movie_img)))
h, w, c = io.imread(path.join(FULL_PATH, movie_img,
img_list[0]))[:, :, :].shape
h = 521
w = 941
# 保存用ディレクトリの作成
save_path = path.join(ROOT_PATH, 'examples', 'result_acrobat_s01',
args.model.split('.')[0])
if not path.exists(save_path):
os.makedirs(save_path)
for i in tqdm(range(0, len(img_list) - 1, 2)):
img1_path = path.join(FULL_PATH, movie_img, img_list[i])
img2_path = path.join(FULL_PATH, movie_img, img_list[i + 2])
data = np.zeros((2, h, w, c)).astype(np.uint8)
data[0] = io.imread(img1_path)[:h, :w, :]
data[1] = io.imread(img2_path)[:h, :w, :]
data = data.transpose(0, 3, 1, 2).reshape(1, 2, c, h, w).astype(
np.float32) / 255.0
data = cp.array(data)
with chainer.using_config('train', False):
with chainer.using_config('enable_backprop', False):
data = model.generator(data)
data = chainer.cuda.to_cpu(data.data) * 255
data = np.clip(data, 0, 255)
data = data.astype(np.uint8)
data = data.reshape(c, h, w).transpose(1, 2, 0)
name, _ = img_list[i].split('.')
img1 = io.imread(img1_path)[:h, :w, :]
io.imsave(
path.join(save_path, '{:}_fi_{:04d}.bmp'.format(movie_img, i + 1)),
img1)
io.imsave(
path.join(save_path, '{:}_fi_{:04d}.bmp'.format(movie_img, i + 2)),
data)
def main():
'''
main function, start point
'''
# 引数関連
parser = argparse.ArgumentParser()
parser.add_argument('--batchsize',
'-b',
type=int,
default=128,
help='Number of images in each mini-batch')
parser.add_argument('--learnrate',
'-l',
type=float,
default=0.001,
help='Learning rate for SGD')
parser.add_argument('--epoch',
'-e',
type=int,
default=300,
help='Number of sweeps over the dataset to train')
parser.add_argument('--gpu',
'-g',
type=int,
default=0,
help='GPU ID (negative value indicates CPU)')
parser.add_argument('--resume',
'-r',
default='',
help='Resume the training from snapshot')
parser.add_argument('--iter_parallel',
action='store_true',
default=False,
help='filter(kernel) sizes')
parser.add_argument('--opt',
'-o',
default='sgd',
help='Resume the training from snapshot')
parser.add_argument('--fsize',
'-f',
default=3,
type=int,
help='Resume the training from snapshot')
parser.add_argument('--ch',
'-c',
default=1,
type=int,
help='Resume the training from snapshot')
args = parser.parse_args()
# parameter出力
print("-=Learning Parameter=-")
print("# Max Epochs: {}".format(args.epoch))
print("# Batch Size: {}".format(args.batchsize))
print("# Learning Rate: {}".format(args.learnrate))
print('# Train Dataet: General 100')
print('# Test Dataet: Set 14')
if args.iter_parallel:
print("# Data Iters that loads in Parallel")
print("\n")
# 保存ディレクトリ
# save didrectory
model_dir_name = 'AESR_opt_{}_fs_{}_ch_{}'.format(args.opt, args.fsize,
args.ch)
outdir = path.join(ROOT_PATH, 'results', model_dir_name)
if not path.exists(outdir):
os.makedirs(outdir)
with open(path.join(outdir, 'arg_param.txt'), 'w') as f:
for k, v in args.__dict__.items():
f.write('{}:{}\n'.format(k, v))
print('# loading dataet(General100, Set14) ...')
train, test = load_dataset()
# prepare model
model = N.GenEvaluator(N.AESR(f_size=args.fsize, ch=args.ch))
if args.gpu >= 0:
chainer.cuda.get_device_from_id(args.gpu).use()
model.to_gpu()
# setup optimizer
if args.opt == 'sgd':
optimizer = chainer.optimizers.MomentumSGD(lr=args.learnrate,
momentum=0.9)
optimizer.setup(model)
optimizer.add_hook(chainer.optimizer.WeightDecay(0.0001))
optimizer.add_hook(chainer.optimizer.GradientClipping(0.1))
elif args.opt == 'adam':
optimizer = chainer.optimizers.Adam(alpha=args.learnrate)
optimizer.setup(model)
optimizer.add_hook(chainer.optimizer.WeightDecay(0.0001))
# setup iter
if args.iter_parallel:
train_iter = chainer.iterators.MultiprocessIterator(train,
args.batchsize,
n_processes=8)
test_iter = chainer.iterators.MultiprocessIterator(test,
args.batchsize,
repeat=False,
shuffle=False,
n_processes=8)
else:
train_iter = chainer.iterators.SerialIterator(train, args.batchsize)
test_iter = chainer.iterators.SerialIterator(test,
args.batchsize,
repeat=False,
shuffle=False)
# setup trainer
updater = training.StandardUpdater(train_iter, optimizer, device=args.gpu)
trainer = training.Trainer(updater, (args.epoch, 'epoch'), out=outdir)
# eval test data
trainer.extend(extensions.Evaluator(test_iter, model, device=args.gpu))
# dump loss graph
trainer.extend(extensions.dump_graph('main/loss'))
# lr shift
if args.opt == 'sgd':
trainer.extend(extensions.ExponentialShift("lr", 0.1),
trigger=(100, 'epoch'))
elif args.opt == 'adam':
trainer.extend(extensions.ExponentialShift("alpha", 0.1),
trigger=(100, 'epoch'))
# save snapshot
trainer.extend(extensions.snapshot(), trigger=(10, 'epoch'))
trainer.extend(extensions.snapshot_object(
model, 'model_snapshot_{.updater.epoch}'),
trigger=(10, 'epoch'))
# log report
trainer.extend(extensions.LogReport())
trainer.extend(extensions.observe_lr(), trigger=(1, 'epoch'))
# plot loss graph
trainer.extend(
extensions.PlotReport(['main/loss', 'validation/main/loss'],
'epoch',
file_name='loss.png'))
# plot acc graph
trainer.extend(
extensions.PlotReport(['main/PSNR', 'validation/main/PSNR'],
'epoch',
file_name='PSNR.png'))
# print info
trainer.extend(
extensions.PrintReport([
'epoch', 'main/loss', 'validation/main/loss', 'main/PSNR',
'validation/main/PSNR', 'lr', 'elapsed_time'
]))
# print progbar
trainer.extend(extensions.ProgressBar())
trainer.run()
# save final model
model_outdir = path.join(ROOT_PATH, 'models', model_dir_name)
if not path.exists(model_outdir):
os.makedirs(model_outdir)
model_name = 'AESR_opt_{}_ch_{}_fsize_{}.npz'.format(
args.opt, args.ch, args.fsize)
chainer.serializers.save_npz(path.join(model_outdir, model_name), model)
model_parameter = {
'name': 'AESR',
'parameter': {
'f_size': args.fsize,
'ch': args.ch
}
}
with open(path.join(model_outdir, 'model_parameter.json'), 'w') as f:
json.dump(model_parameter, f)
def main():
'''
main function, start point
'''
# 引数関連
parser = argparse.ArgumentParser()
parser.add_argument('--batchsize',
'-b',
type=int,
default=128,
help='Number of images in each mini-batch')
parser.add_argument('--learnrate',
'-l',
type=float,
default=0.001,
help='Learning rate for SGD')
parser.add_argument('--epoch',
'-e',
type=int,
default=100,
help='Number of sweeps over the dataset to train')
parser.add_argument('--gpu0',
'-g',
type=int,
default=0,
help='GPU1 ID (negative value indicates CPU)')
parser.add_argument('--gpu1',
'-G',
type=int,
default=2,
help='GPU2 ID (negative value indicates CPU)')
parser.add_argument('--resume',
'-r',
default='',
help='Resume the training from snapshot')
parser.add_argument('--iter_parallel',
'-p',
action='store_true',
default=False,
help='loading dataset from disk')
parser.add_argument('--opt',
'-o',
type=str,
choices=('adam', 'sgd'),
default='adam')
parser.add_argument('--fsize', '-f', type=int, default=5)
parser.add_argument('--ch', '-c', type=int, default=4)
args = parser.parse_args()
# parameter出力
print("-=Learning Parameter=-")
print("# Max Epochs: {}".format(args.epoch))
print("# Batch Size: {}".format(args.batchsize))
print("# Learning Rate: {}".format(args.learnrate))
print("# Optimizer Method: {}".format(args.opt))
print("# Filter Size: {}".format(args.fsize))
print("# Channel Scale: {}".format(args.ch))
print('# Train Dataet: General 100')
if args.iter_parallel:
print("# Data Iters that loads in Parallel")
print("\n")
# 保存ディレクトリ
# make result dir
network_name = 'AEFINetConcat'
model_name = 'AEFINet_Test_opt_{}_ch_{}_fsize_{}'.format(
args.opt, args.ch, args.fsize)
outdir = path.join(ROOT_PATH, 'results', 'FI', 'AEFINet', model_name)
util.make_result_dir(args, outdir)
#loading dataset
if args.iter_parallel:
train = datasets.SequenceDataset(
dataset='UCF101_train_size64_frame3_group10_max100_p')
test = datasets.SequenceDataset(
dataset='UCF101_test_size64_frame3_group25_max5_p')
else:
train = datasets.SequenceDatasetOnMem(
dataset='UCF101_train_size64_frame3_group10_max100_p')
test = datasets.SequenceDatasetOnMem(
dataset='UCF101_test_size64_frame3_group25_max5_p')
# prepare model
chainer.cuda.get_device_from_id(args.gpu0).use()
model = N.GenEvaluator(N.AEFINetConcat(f_size=args.fsize, ch=args.ch))
# setup optimizer
if args.opt == 'adam':
optimizer = chainer.optimizers.Adam(alpha=args.learnrate)
elif args.opt == 'sgd':
optimizer = chainer.optimizers.MomentumSGD(lr=args.learnrate,
momentum=0.9)
optimizer.setup(model)
optimizer.add_hook(chainer.optimizer.WeightDecay(0.0001))
# setup iter
if args.iter_parallel:
train_iter = chainer.iterators.MultiprocessIterator(train,
args.batchsize,
n_processes=8)
test_iter = chainer.iterators.MultiprocessIterator(test,
args.batchsize,
repeat=False,
shuffle=False,
n_processes=8)
else:
train_iter = chainer.iterators.SerialIterator(train, args.batchsize)
test_iter = chainer.iterators.SerialIterator(test,
args.batchsize,
repeat=False,
shuffle=False)
# setup trainer
updater = training.ParallelUpdater(
train_iter,
optimizer,
devices={
'main': args.gpu0,
'second': args.gpu1
},
)
trainer = training.Trainer(updater, (args.epoch, 'epoch'), out=outdir)
# # eval test data
trainer.extend(extensions.Evaluator(test_iter, model, device=args.gpu0))
# dump loss graph
trainer.extend(extensions.dump_graph('main/loss'))
# lr shift
if args.opt == 'sgd':
trainer.extend(extensions.ExponentialShift("lr", 0.1),
trigger=(100, 'epoch'))
elif args.opt == 'adam':
trainer.extend(extensions.ExponentialShift("alpha", 0.1),
trigger=(100, 'epoch'))
# save snapshot
trainer.extend(extensions.snapshot(), trigger=(10, 'epoch'))
trainer.extend(extensions.snapshot_object(
model, 'model_snapshot_{.updater.epoch}'),
trigger=(10, 'epoch'))
# log report
trainer.extend(extensions.LogReport())
trainer.extend(extensions.observe_lr(), trigger=(1, 'epoch'))
# plot loss graph
trainer.extend(
extensions.PlotReport(['main/loss', 'validation/main/loss'],
'epoch',
file_name='loss.png'))
# plot acc graph
trainer.extend(
extensions.PlotReport(['main/PSNR', 'validation/main/PSNR'],
'epoch',
file_name='PSNR.png'))
# print info
trainer.extend(
extensions.PrintReport([
'epoch', 'main/loss', 'validation/main/loss', 'main/PSNR',
'validation/main/PSNR', 'lr', 'elapsed_time'
]))
# print progbar
trainer.extend(extensions.ProgressBar())
# [ChainerUI] enable to send commands from ChainerUI
trainer.extend(CommandsExtension())
# [ChainerUI] save 'args' to show experimental conditions
save_args(args, outdir)
if args.resume:
# Resume from a snapshot
chainer.serializers.load_npz(args.resume, trainer)
# start train
trainer.run()
# save final model
util.save_trained_model(model_name,
model,
network_name,
f_size=args.fsize,
ch=args.ch)
