def display(tfrecord_dir):
print('Loading dataset "%s"' % tfrecord_dir)
tfutil.init_tf({'gpu_options.allow_growth': True})
dset = dataset.TFRecordDataset(tfrecord_dir,
max_label_size='full',
repeat=False,
shuffle_mb=0)
tfutil.init_uninited_vars()
idx = 0
while True:
try:
images, labels = dset.get_minibatch_np(1)
except tf.errors.OutOfRangeError:
break
if idx == 0:
print('Displaying images')
import cv2 # pip install opencv-python
cv2.namedWindow('dataset_tool')
print('Press SPACE or ENTER to advance, ESC to exit')
print('\nidx = %-8d\nlabel = %s' % (idx, labels[0].tolist()))
cv2.imshow('dataset_tool', images[0].transpose(
1, 2, 0)[:, :, ::-1]) # CHW => HWC, RGB => BGR
idx += 1
if cv2.waitKey() == 27:
break
print('\nDisplayed %d images.' % idx)
def extract(tfrecord_dir, output_dir):
print('Loading dataset "%s"' % tfrecord_dir)
tfutil.init_tf({'gpu_options.allow_growth': True})
dset = dataset.TFRecordDataset(tfrecord_dir,
max_label_size=0,
repeat=False,
shuffle_mb=0)
tfutil.init_uninited_vars()
print('Extracting images to "%s"' % output_dir)
if not os.path.isdir(output_dir):
os.makedirs(output_dir)
idx = 0
while True:
if idx % 10 == 0:
print('%d\r' % idx, end='', flush=True)
try:
images, labels = dset.get_minibatch_np(1)
except tf.errors.OutOfRangeError:
break
if images.shape[1] == 1:
img = PIL.Image.fromarray(images[0][0], 'L')
else:
img = PIL.Image.fromarray(images[0].transpose(1, 2, 0), 'RGB')
img.save(os.path.join(output_dir, 'img%08d.png' % idx))
idx += 1
print('Extracted %d images.' % idx)
def compare(tfrecord_dir_a, tfrecord_dir_b, ignore_labels):
max_label_size = 0 if ignore_labels else 'full'
print('Loading dataset "%s"' % tfrecord_dir_a)
tfutil.init_tf({'gpu_options.allow_growth': True})
dset_a = dataset.TFRecordDataset(
tfrecord_dir_a,
max_label_size=max_label_size,
repeat=False,
shuffle_mb=0)
print('Loading dataset "%s"' % tfrecord_dir_b)
dset_b = dataset.TFRecordDataset(
tfrecord_dir_b,
max_label_size=max_label_size,
repeat=False,
shuffle_mb=0)
tfutil.init_uninited_vars()
print('Comparing datasets')
idx = 0
identical_images = 0
identical_labels = 0
while True:
if idx % 100 == 0:
print('%d\r' % idx, end='', flush=True)
try:
images_a, labels_a = dset_a.get_minibatch_np(1)
except tf.errors.OutOfRangeError:
images_a, labels_a = None, None
try:
images_b, labels_b = dset_b.get_minibatch_np(1)
except tf.errors.OutOfRangeError:
images_b, labels_b = None, None
if images_a is None or images_b is None:
if images_a is not None or images_b is not None:
print('Datasets contain different number of images')
break
if images_a.shape == images_b.shape and np.all(images_a == images_b):
identical_images += 1
else:
print('Image %d is different' % idx)
if labels_a.shape == labels_b.shape and np.all(labels_a == labels_b):
identical_labels += 1
else:
print('Label %d is different' % idx)
idx += 1
print('Identical images: %d / %d' % (identical_images, idx))
if not ignore_labels:
print('Identical labels: %d / %d' % (identical_labels, idx))
def extract(tfrecord_dir, output_dir):
print('Loading dataset "%s"' % tfrecord_dir)
tfutil.init_tf({'gpu_options.allow_growth': True})
dset = dataset.TFRecordDataset(tfrecord_dir,
max_label_size=0,
repeat=False,
shuffle_mb=0)
tfutil.init_uninited_vars()
### shifting kernel
fc_x = 0.5
fc_y = 0.5
im_size = 128
kernel_loc = 2.*np.pi*fc_x * np.arange(im_size).reshape((1, im_size, 1)) + \
2.*np.pi*fc_y * np.arange(im_size).reshape((im_size, 1, 1))
kernel_cos = np.cos(kernel_loc)
print('Extracting images to "%s"' % output_dir)
if not os.path.isdir(output_dir):
os.makedirs(output_dir)
idx = 0
while True:
if idx % 10 == 0:
print('%d\r' % idx, end='', flush=True)
try:
images, labels = dset.get_minibatch_np(1)
except tf.errors.OutOfRangeError:
break
if images.shape[1] == 1:
img = PIL.Image.fromarray(images[0][0], 'L')
else:
img = PIL.Image.fromarray(images[0].transpose(1, 2, 0), 'RGB')
### shifting the image
#img = np.asarray(img)
#img_t = 2.* (img.astype(np.float32) / 255.) - 1.
#img_sh = img_t * kernel_cos
#img = (img_sh + 1.) / 2. * 255.
#img = np.rint(img).clip(0, 255).astype(np.uint8)
#img = PIL.Image.fromarray(img)
img.save(os.path.join(output_dir, 'img%08d.png' % idx))
idx += 1
print('Extracted %d images.' % idx)
def compare(tfrecord_dir_a, tfrecord_dir_b, ignore_labels):
max_label_size = 0 if ignore_labels else 'full'
print('Loading dataset "%s"' % tfrecord_dir_a)
tfutil.init_tf({'gpu_options.allow_growth': True})
dset_a = dataset.TFRecordDataset(tfrecord_dir_a, max_label_size=max_label_size, repeat=False, shuffle_mb=0)
print('Loading dataset "%s"' % tfrecord_dir_b)
dset_b = dataset.TFRecordDataset(tfrecord_dir_b, max_label_size=max_label_size, repeat=False, shuffle_mb=0)
tfutil.init_uninited_vars()
print('Comparing datasets')
idx = 0
identical_images = 0
identical_labels = 0
while True:
if idx % 100 == 0:
print('%d\r' % idx, end='', flush=True)
try:
images_a, labels_a = dset_a.get_minibatch_np(1)
except tf.errors.OutOfRangeError:
images_a, labels_a = None, None
try:
images_b, labels_b = dset_b.get_minibatch_np(1)
except tf.errors.OutOfRangeError:
images_b, labels_b = None, None
if images_a is None or images_b is None:
if images_a is not None or images_b is not None:
print('Datasets contain different number of images')
break
if images_a.shape == images_b.shape and np.all(images_a == images_b):
identical_images += 1
else:
print('Image %d is different' % idx)
if labels_a.shape == labels_b.shape and np.all(labels_a == labels_b):
identical_labels += 1
else:
print('Label %d is different' % idx)
idx += 1
print('Identical images: %d / %d' % (identical_images, idx))
if not ignore_labels:
print('Identical labels: %d / %d' % (identical_labels, idx))
def display(tfrecord_dir):
print('Loading dataset "%s"' % tfrecord_dir)
tfutil.init_tf({'gpu_options.allow_growth': True})
dset = dataset.TFRecordDataset(tfrecord_dir, max_label_size='full', repeat=False, shuffle_mb=0)
tfutil.init_uninited_vars()
idx = 0
while True:
try:
images, labels = dset.get_minibatch_np(1)
except tf.errors.OutOfRangeError:
break
if idx == 0:
print('Displaying images')
import cv2 # pip install opencv-python
cv2.namedWindow('dataset_tool')
print('Press SPACE or ENTER to advance, ESC to exit')
print('\nidx = %-8d\nlabel = %s' % (idx, labels[0].tolist()))
cv2.imshow('dataset_tool', images[0].transpose(1, 2, 0)[:, :, ::-1]) # CHW => HWC, RGB => BGR
idx += 1
if cv2.waitKey() == 27:
break
print('\nDisplayed %d images.' % idx)
def extract(tfrecord_dir, output_dir):
print('Loading dataset "%s"' % tfrecord_dir)
tfutil.init_tf({'gpu_options.allow_growth': True})
dset = dataset.TFRecordDataset(tfrecord_dir, max_label_size=0, repeat=False, shuffle_mb=0)
tfutil.init_uninited_vars()
print('Extracting images to "%s"' % output_dir)
if not os.path.isdir(output_dir):
os.makedirs(output_dir)
idx = 0
while True:
if idx % 10 == 0:
print('%d\r' % idx, end='', flush=True)
try:
images, labels = dset.get_minibatch_np(1)
except tf.errors.OutOfRangeError:
break
if images.shape[1] == 1:
img = PIL.Image.fromarray(images[0][0], 'L')
else:
img = PIL.Image.fromarray(images[0].transpose(1, 2, 0), 'RGB')
img.save(os.path.join(output_dir, 'img%08d.png' % idx))
idx += 1
print('Extracted %d images.' % idx)
def evaluate_metrics(run_id,
log,
metrics,
num_images,
real_passes,
minibatch_size=None):
metric_class_names = {
'swd': 'metrics.sliced_wasserstein.API',
'fid': 'metrics.frechet_inception_distance.API',
'is': 'metrics.inception_score.API',
'msssim': 'metrics.ms_ssim.API',
}
# Locate training run and initialize logging.
result_subdir = misc.locate_result_subdir(run_id)
snapshot_pkls = misc.list_network_pkls(result_subdir, include_final=False)
assert len(snapshot_pkls) >= 1
log_file = os.path.join(result_subdir, log)
print('Logging output to', log_file)
misc.set_output_log_file(log_file)
# Initialize dataset and select minibatch size.
dataset_obj, mirror_augment = misc.load_dataset_for_previous_run(
result_subdir, verbose=True, shuffle_mb=0)
if minibatch_size is None:
minibatch_size = np.clip(8192 // dataset_obj.shape[1], 4, 256)
# Initialize metrics.
metric_objs = []
for name in metrics:
class_name = metric_class_names.get(name, name)
print('Initializing %s...' % class_name)
class_def = tfutil.import_obj(class_name)
image_shape = [3] + dataset_obj.shape[1:]
obj = class_def(num_images=num_images,
image_shape=image_shape,
image_dtype=np.uint8,
minibatch_size=minibatch_size)
tfutil.init_uninited_vars()
mode = 'warmup'
obj.begin(mode)
for idx in range(10):
obj.feed(
mode,
np.random.randint(0,
256,
size=[minibatch_size] + image_shape,
dtype=np.uint8))
obj.end(mode)
metric_objs.append(obj)
# Print table header.
print()
print('%-10s%-12s' % ('Snapshot', 'Time_eval'), end='')
for obj in metric_objs:
for name, fmt in zip(obj.get_metric_names(),
obj.get_metric_formatting()):
print('%-*s' % (len(fmt % 0), name), end='')
print()
print('%-10s%-12s' % ('---', '---'), end='')
for obj in metric_objs:
for fmt in obj.get_metric_formatting():
print('%-*s' % (len(fmt % 0), '---'), end='')
print()
# Feed in reals.
for title, mode in [('Reals', 'reals'), ('Reals2', 'fakes')][:real_passes]:
print('%-10s' % title, end='')
time_begin = time.time()
labels = np.zeros([num_images, dataset_obj.label_size],
dtype=np.float32)
[obj.begin(mode) for obj in metric_objs]
for begin in range(0, num_images, minibatch_size):
end = min(begin + minibatch_size, num_images)
images, labels[begin:end] = dataset_obj.get_minibatch_np(end -
begin)
if mirror_augment:
images = misc.apply_mirror_augment(images)
if images.shape[1] == 1:
images = np.tile(images, [1, 3, 1, 1]) # grayscale => RGB
[obj.feed(mode, images) for obj in metric_objs]
results = [obj.end(mode) for obj in metric_objs]
print('%-12s' % misc.format_time(time.time() - time_begin), end='')
for obj, vals in zip(metric_objs, results):
for val, fmt in zip(vals, obj.get_metric_formatting()):
print(fmt % val, end='')
print()
# Evaluate each network snapshot.
for snapshot_idx, snapshot_pkl in enumerate(reversed(snapshot_pkls)):
prefix = 'network-snapshot-'
postfix = '.pkl'
snapshot_name = os.path.basename(snapshot_pkl)
assert snapshot_name.startswith(prefix) and snapshot_name.endswith(
postfix)
snapshot_kimg = int(snapshot_name[len(prefix):-len(postfix)])
print('%-10d' % snapshot_kimg, end='')
mode = 'fakes'
[obj.begin(mode) for obj in metric_objs]
time_begin = time.time()
with tf.Graph().as_default(), tfutil.create_session(
config.tf_config).as_default():
G, D, Gs = misc.load_pkl(snapshot_pkl)
for begin in range(0, num_images, minibatch_size):
end = min(begin + minibatch_size, num_images)
latents = misc.random_latents(end - begin, Gs)
images = Gs.run(latents,
labels[begin:end],
num_gpus=config.num_gpus,
out_mul=127.5,
out_add=127.5,
out_dtype=np.uint8)
if images.shape[1] == 1:
images = np.tile(images, [1, 3, 1, 1]) # grayscale => RGB
[obj.feed(mode, images) for obj in metric_objs]
results = [obj.end(mode) for obj in metric_objs]
print('%-12s' % misc.format_time(time.time() - time_begin), end='')
for obj, vals in zip(metric_objs, results):
for val, fmt in zip(vals, obj.get_metric_formatting()):
print(fmt % val, end='')
print()
print()
def evaluate_metrics_swd_distributions_training_trad_prog(
run_id,
network_dir_conv,
network_dir_prog,
log,
metrics,
num_images_per_group,
num_groups,
real_passes,
minibatch_size=None):
metric_class_names = {
'swd_distri_training_trad_prog':
'metrics.swd_distributions_training_trad_prog.API',
}
# Locate training run and initialize logging.
result_subdir = misc.locate_result_subdir(run_id)
log_file = os.path.join(result_subdir, log)
print('Logging output to', log_file)
misc.set_output_log_file(log_file)
# Initialize dataset and select minibatch size.
dataset_obj, mirror_augment = misc.load_dataset_for_previous_run(
result_subdir, verbose=True, shuffle_mb=0)
# Initialize metrics.
metric_objs = []
for name in metrics:
class_name = metric_class_names.get(name, name)
print('Initializing %s...' % class_name)
class_def = tfutil.import_obj(class_name)
image_shape = [3] + dataset_obj.shape[1:]
obj = class_def(image_shape=image_shape, image_dtype=np.uint8)
tfutil.init_uninited_vars()
metric_objs.append(obj)
mode = 'fakes'
[obj.begin(mode) for obj in metric_objs]
images_real, labels = dataset_obj.get_minibatch_np(
num_groups * num_images_per_group)
with tf.Graph().as_default(), tfutil.create_session(
config.tf_config).as_default():
G, D, Gs = misc.load_pkl(network_dir_conv)
#G, D, Gs = pickle.load(file)
latents = misc.random_latents(num_groups * num_images_per_group,
Gs)
images = images_real
for k in range(
10
): # because Gs can not generate lots of (>3000 around) images at one time. Make sure /10 = int
nn = int(num_groups * num_images_per_group / 10)
images_fake = Gs.run(latents[k * nn:(k + 1) * nn],
labels[k * nn:(k + 1) * nn],
num_gpus=config.num_gpus,
out_mul=127.5,
out_add=127.5,
out_dtype=np.uint8)
images = np.concatenate((images, images_fake), axis=0)
with tf.Graph().as_default(), tfutil.create_session(
config.tf_config).as_default():
G, D, Gs = misc.load_pkl(network_dir_prog)
# G, D, Gs = pickle.load(file)
latents = misc.random_latents(num_groups * num_images_per_group,
Gs)
for k in range(
10
): # because Gs can not generate lots of (>3000 around) images at one time. Make sure /10 = int
nn = int(num_groups * num_images_per_group / 10)
images_fake = Gs.run(latents[k * nn:(k + 1) * nn],
labels[k * nn:(k + 1) * nn],
num_gpus=config.num_gpus,
out_mul=127.5,
out_add=127.5,
out_dtype=np.uint8)
images = np.concatenate((images, images_fake), axis=0)
if images.shape[1] == 1:
images = np.tile(images, [1, 3, 1, 1]) # grayscale => RGB
[
obj.feed(mode, images, num_images_per_group, num_groups,
result_subdir) for obj in metric_objs
]
def evaluate_metrics_swd_distributions(run_id,
log,
metrics,
num_images_per_group,
num_groups,
real_passes,
minibatch_size=None):
metric_class_names = {
'swd_distri': 'metrics.swd_distributions.API',
}
# Locate training run and initialize logging.
result_subdir = misc.locate_result_subdir(run_id)
snapshot_pkls = misc.list_network_pkls(result_subdir, include_final=False)
assert len(snapshot_pkls) >= 1
log_file = os.path.join(result_subdir, log)
print('Logging output to', log_file)
misc.set_output_log_file(log_file)
# Initialize dataset and select minibatch size.
dataset_obj, mirror_augment = misc.load_dataset_for_previous_run(
result_subdir, verbose=True, shuffle_mb=0)
# Initialize metrics.
metric_objs = []
for name in metrics:
class_name = metric_class_names.get(name, name)
print('Initializing %s...' % class_name)
class_def = tfutil.import_obj(class_name)
image_shape = [3] + dataset_obj.shape[1:]
obj = class_def(image_shape=image_shape, image_dtype=np.uint8)
tfutil.init_uninited_vars()
metric_objs.append(obj)
# Evaluate each network snapshot.
for snapshot_idx, snapshot_pkl in enumerate(reversed(snapshot_pkls)):
prefix = 'network-snapshot-'
postfix = '.pkl'
snapshot_name = os.path.basename(snapshot_pkl)
assert snapshot_name.startswith(prefix) and snapshot_name.endswith(
postfix)
snapshot_kimg = int(snapshot_name[len(prefix):-len(postfix)])
print('%-10d' % snapshot_kimg, end='')
mode = 'fakes'
[obj.begin(mode) for obj in metric_objs]
images_real, labels = dataset_obj.get_minibatch_np(
num_groups * num_images_per_group)
with tf.Graph().as_default(), tfutil.create_session(
config.tf_config).as_default():
G, D, Gs = misc.load_pkl(snapshot_pkl)
latents = misc.random_latents(num_groups * num_images_per_group,
Gs)
images = images_real
for k in range(
10
): # because Gs can not generate lots of (>3000 around) images at one time. Make sure /10 = int
nn = int(num_groups * num_images_per_group / 10)
images_fake = Gs.run(latents[k * nn:(k + 1) * nn],
labels[k * nn:(k + 1) * nn],
num_gpus=config.num_gpus,
out_mul=127.5,
out_add=127.5,
out_dtype=np.uint8)
images = np.concatenate((images, images_fake), axis=0)
if images.shape[1] == 1:
images = np.tile(images, [1, 3, 1, 1]) # grayscale => RGB
[
obj.feed(mode, images, num_images_per_group, num_groups,
snapshot_kimg, result_subdir) for obj in metric_objs
]
def evaluate_metrics(run_id, log, metrics, num_images, real_passes, minibatch_size=None):
metric_class_names = {
'swd': 'metrics.sliced_wasserstein.API',
'fid': 'metrics.frechet_inception_distance.API',
'is': 'metrics.inception_score.API',
'msssim': 'metrics.ms_ssim.API',
}
# Locate training run and initialize logging.
result_subdir = misc.locate_result_subdir(run_id)
snapshot_pkls = misc.list_network_pkls(result_subdir, include_final=False)
assert len(snapshot_pkls) >= 1
log_file = os.path.join(result_subdir, log)
print('Logging output to', log_file)
misc.set_output_log_file(log_file)
# Initialize dataset and select minibatch size.
dataset_obj, mirror_augment = misc.load_dataset_for_previous_run(result_subdir, verbose=True, shuffle_mb=0)
if minibatch_size is None:
minibatch_size = np.clip(8192 // dataset_obj.shape[1], 4, 256)
# Initialize metrics.
metric_objs = []
for name in metrics:
class_name = metric_class_names.get(name, name)
print('Initializing %s...' % class_name)
class_def = tfutil.import_obj(class_name)
image_shape = [3] + dataset_obj.shape[1:]
obj = class_def(num_images=num_images, image_shape=image_shape, image_dtype=np.uint8, minibatch_size=minibatch_size)
tfutil.init_uninited_vars()
mode = 'warmup'
obj.begin(mode)
for idx in range(10):
obj.feed(mode, np.random.randint(0, 256, size=[minibatch_size]+image_shape, dtype=np.uint8))
obj.end(mode)
metric_objs.append(obj)
# Print table header.
print()
print('%-10s%-12s' % ('Snapshot', 'Time_eval'), end='')
for obj in metric_objs:
for name, fmt in zip(obj.get_metric_names(), obj.get_metric_formatting()):
print('%-*s' % (len(fmt % 0), name), end='')
print()
print('%-10s%-12s' % ('---', '---'), end='')
for obj in metric_objs:
for fmt in obj.get_metric_formatting():
print('%-*s' % (len(fmt % 0), '---'), end='')
print()
# Feed in reals.
for title, mode in [('Reals', 'reals'), ('Reals2', 'fakes')][:real_passes]:
print('%-10s' % title, end='')
time_begin = time.time()
labels = np.zeros([num_images, dataset_obj.label_size], dtype=np.float32)
[obj.begin(mode) for obj in metric_objs]
for begin in range(0, num_images, minibatch_size):
end = min(begin + minibatch_size, num_images)
images, labels[begin:end] = dataset_obj.get_minibatch_np(end - begin)
if mirror_augment:
images = misc.apply_mirror_augment(images)
if images.shape[1] == 1:
images = np.tile(images, [1, 3, 1, 1]) # grayscale => RGB
[obj.feed(mode, images) for obj in metric_objs]
results = [obj.end(mode) for obj in metric_objs]
print('%-12s' % misc.format_time(time.time() - time_begin), end='')
for obj, vals in zip(metric_objs, results):
for val, fmt in zip(vals, obj.get_metric_formatting()):
print(fmt % val, end='')
print()
# Evaluate each network snapshot.
for snapshot_idx, snapshot_pkl in enumerate(reversed(snapshot_pkls)):
prefix = 'network-snapshot-'; postfix = '.pkl'
snapshot_name = os.path.basename(snapshot_pkl)
assert snapshot_name.startswith(prefix) and snapshot_name.endswith(postfix)
snapshot_kimg = int(snapshot_name[len(prefix) : -len(postfix)])
print('%-10d' % snapshot_kimg, end='')
mode ='fakes'
[obj.begin(mode) for obj in metric_objs]
time_begin = time.time()
with tf.Graph().as_default(), tfutil.create_session(config.tf_config).as_default():
G, D, Gs = misc.load_pkl(snapshot_pkl)
for begin in range(0, num_images, minibatch_size):
end = min(begin + minibatch_size, num_images)
latents = misc.random_latents(end - begin, Gs)
images = Gs.run(latents, labels[begin:end], num_gpus=config.num_gpus, out_mul=127.5, out_add=127.5, out_dtype=np.uint8)
if images.shape[1] == 1:
images = np.tile(images, [1, 3, 1, 1]) # grayscale => RGB
[obj.feed(mode, images) for obj in metric_objs]
results = [obj.end(mode) for obj in metric_objs]
print('%-12s' % misc.format_time(time.time() - time_begin), end='')
for obj, vals in zip(metric_objs, results):
for val, fmt in zip(vals, obj.get_metric_formatting()):
print(fmt % val, end='')
print()
print()