def main():
image = tf.placeholder(dtype=tf.float32, shape=[1, None, None, 3], name='image')
net, end_points = deeplabv3(image, num_classes=args.num_classes, depth=args.num_layers, is_training=False)
raw_output = end_points['resnet_v1_{}/logits'.format(args.num_layers)]
with tf.variable_scope('inference_out'):
seg_pred = tf.argmax(raw_output, axis=3)
# Set up tf session and initialize variables.
config = tf.ConfigProto()
config.gpu_options.allow_growth = True
graph = tf.get_default_graph()
input_graph_def = graph.as_graph_def()
with tf.Session(config=config) as sess:
image_data = np.array(Image.open('./datasets/IDImage/JPGImage/id00001.jpg'))
image_data = np.expand_dims(image_data,0)
sess.run(tf.global_variables_initializer())
sess.run(tf.local_variables_initializer())
restore_var = [v for v in tf.global_variables()]
loader = tf.train.Saver(var_list=restore_var)
load(loader, sess, args.snapshot_dir)
run_ops = [seg_pred];output_node_names = ['inference_out/ArgMax']
sess.run(run_ops, feed_dict={image: image_data,})
output_graph_def = graph_util.convert_variables_to_constants(sess, input_graph_def,output_node_names)
with tf.gfile.GFile(os.path.join(pb_path, "model.pb"), "wb") as f:
f.write(output_graph_def.SerializeToString())
def test_setup(self):
# Create queue coordinator.
num_layers = 50
self.coord = tf.train.Coordinator()
# Load reader
with tf.name_scope("create_inputs"):
reader = ImageReader(
self.conf.data_dir,
self.conf.valid_data_list,
None, # the images have different sizes
False, # no data-aug
False, # no data-aug
self.conf.ignore_label,
IMG_MEAN,
self.coord)
image, label = reader.image, reader.label # [h, w, 3 or 1]
# Add one batch dimension [1, h, w, 3 or 1]
self.image_batch, self.label_batch = tf.expand_dims(
image, dim=0), tf.expand_dims(label, dim=0)
# Create network
net, end_points = deeplabv3(self.image_batch,
num_classes=self.conf.num_classes,
depth=num_layers,
is_training=True)
raw_output = end_points['resnet{}/logits'.format(num_layers)]
# predictions
#raw_output = net.o # [batch_size, 41, 41, 21]
raw_output = tf.image.resize_bilinear(
raw_output,
tf.shape(self.image_batch)[1:3, ])
raw_output = tf.argmax(raw_output, axis=3)
pred = tf.expand_dims(raw_output, dim=3)
self.pred = tf.reshape(pred, [
-1,
])
# labels
gt = tf.reshape(self.label_batch, [
-1,
])
# Ignoring all labels greater than or equal to n_classes.
temp = tf.less_equal(gt, self.conf.num_classes - 1)
weights = tf.cast(temp, tf.int32)
# fix for tf 1.3.0
gt = tf.where(temp, gt, tf.cast(temp, tf.uint8))
# Pixel accuracy
self.accu, self.accu_update_op = tf.contrib.metrics.streaming_accuracy(
self.pred, gt, weights=weights)
# mIoU
self.mIoU, self.mIou_update_op = tf.contrib.metrics.streaming_mean_iou(
self.pred, gt, num_classes=self.conf.num_classes, weights=weights)
# Loader for loading the checkpoint
self.loader = tf.train.Saver(var_list=tf.global_variables())
def main():
"""Create the model and start the training."""
tf.set_random_seed(args.random_seed)
# Create queue coordinator.
coord = tf.train.Coordinator()
image_batch, label_batch = read_data(is_training=args.is_training,
split_name=args.split_name)
# Create network.
net, end_points = deeplabv3(image_batch,
num_classes=args.num_classes,
depth=args.num_layers,
is_training=False)
# For a small batch size, it is better to keep
# the statistics of the BN layers (running means and variances)
# frozen, and to not update the values provided by the pre-trained model.
# If is_training=True, the statistics will be updated during the training.
# Note that is_training=False still updates BN parameters gamma (scale) and beta (offset)
# if they are presented in var_list of the optimiser definition.
# Which variables to load. Running means and variances are not trainable,
# thus all_variables() should be restored.
restore_var = [
v for v in tf.global_variables()
if 'fc' not in v.name or not args.not_restore_last
]
# Predictions.
raw_output = end_points['resnet_v1_{}/logits'.format(args.num_layers)]
# Predictions: ignoring all predictions with labels greater or equal than n_classes
# nh, nw = tf.shape(image_batch)[1], tf.shape(image_batch)[2]
# seg_logits = tf.image.resize_bilinear(raw_output, [nh, nw])
seg_pred = tf.argmax(raw_output, axis=3)
to_save = seg_pred
seg_pred = tf.expand_dims(seg_pred, 3)
seg_pred = tf.reshape(seg_pred, [
-1,
])
seg_gt = tf.cast(label_batch, tf.int32)
seg_gt = tf.reshape(seg_gt, [
-1,
])
mask = seg_gt <= args.num_classes - 1
seg_pred = tf.boolean_mask(seg_pred, mask)
seg_gt = tf.boolean_mask(seg_gt, mask)
mean_iou, update_mean_iou = streaming_mean_iou(
seg_pred, seg_gt, num_classes=args.num_classes)
# Set up tf session and initialize variables.
config = tf.ConfigProto()
config.gpu_options.allow_growth = True
sess = tf.Session(config=config)
sess.run(tf.global_variables_initializer())
sess.run(tf.local_variables_initializer())
# Load variables if the checkpoint is provided.
if args.ckpt > 0 or args.restore_from is not None:
loader = tf.train.Saver(var_list=restore_var)
load(loader, sess, args.snapshot_dir)
# Start queue threads.
threads = tf.train.start_queue_runners(coord=coord, sess=sess)
tf.get_default_graph().finalize()
mean_ious = {}
sum = 0
for step in range(125):
start_time = time.time()
score_map, picture, mean_iou_float, _ = sess.run(
[raw_output, to_save, mean_iou, update_mean_iou])
picture = np.squeeze(picture)
with open('./datasets/IDImage/{}.txt'.format(args.split_name),
'r') as f:
lines = f.readlines()
output_name = lines[step].split('/')[-1][:-1]
mean_ious['{}'.format(output_name.split('.')[0])] = mean_iou_float
save_path = os.path.join(args.output_path, output_name)
pyplot.imsave(save_path, picture)
duration = time.time() - start_time
sys.stdout.write(
'step {:d}, mean_iou: {:.6f}({:.3f} sec/step)\n'.format(
step, mean_iou_float, duration))
sys.stdout.flush()
if coord.should_stop():
coord.request_stop()
coord.join(threads)
for item in mean_ious.values():
sum += item
print('===============================')
print(sum / (NUM_VAL - 1))
print('===============================')
def train_setup(self, reuse=False):
tf.set_random_seed(self.conf.random_seed)
num_layers = 50 #-----------------------------------------------------------------------------------------
# Create queue coordinator.
self.coord = tf.train.Coordinator()
self.n_gpu = self.conf.n_gpu
# Input size
self.input_size = (self.conf.input_height, self.conf.input_width)
j_step = 0
with tf.name_scope("create_inputs"):
reader = ImageReader(self.conf.data_dir, self.conf.data_list,
self.input_size, self.conf.random_scale,
self.conf.random_mirror,
self.conf.ignore_label, IMG_MEAN, self.coord)
# print "1"*22
# print reader
image_data, image_label = reader.dequeue(self.conf.batch_size)
self.image_data = image_data
if tf.__version__.startswith('1.'):
split_train_data_node = tf.split(image_data, self.n_gpu)
split_train_labels_node = tf.split(image_label, self.n_gpu)
else:
split_train_data_node = tf.split(0, self.n_gpu, image_data)
split_train_labels_node = tf.split(0, self.n_gpu, image_label)
with tf.variable_scope(tf.get_variable_scope()):
all_loss = []
for device_index, (i, self.image_batch,
self.label_batch) in enumerate(
zip([1], split_train_data_node,
split_train_labels_node)):
with tf.device('/gpu:%d' % i):
#print i
with tf.name_scope('%s_%d' % ("gpu", i)) as scope:
if j_step == 0:
j_step = 1
pass
else:
reuse = True
# net = DeepLab_v2_Network(self.image_batch, num_classes=self.conf.num_classes,
# is_training=self.conf.is_training ,reuse=reuse)
net, end_points = deeplabv3(
self.image_batch,
num_classes=self.conf.num_classes,
depth=num_layers,
is_training=True,
reuse=reuse)
self.raw_output = end_points[
'gpu_{}/resnet{}/logits'.format(i, num_layers)]
# Network raw output
# [batch_size, 41, 41, 21]
output_size = (self.raw_output.shape[1].value,
self.raw_output.shape[2].value)
label_proc = prepare_label(
self.label_batch,
output_size,
num_classes=self.conf.num_classes,
one_hot=False) # [batch_size, 41, 41]
raw_gt = tf.reshape(label_proc, [
-1,
])
indices = tf.squeeze(
tf.where(
tf.less_equal(raw_gt,
self.conf.num_classes - 1)), 1)
gt = tf.cast(tf.gather(raw_gt, indices), tf.int32)
raw_prediction = tf.reshape(
self.raw_output, [-1, self.conf.num_classes])
# print raw_prediction
# print gt
prediction = raw_prediction
# prediction = tf.expand_dims(raw_prediction, 3)
# prediction = tl.act.pixel_wise_softmax(prediction)
# print prediction
# print label_proc
# loss = 1 - tl.cost.dice_coe(prediction, label_proc, axis=[1, 2, 3, 4])
loss = tf.nn.sparse_softmax_cross_entropy_with_logits(
logits=prediction, labels=gt)
l2_losses = [
self.conf.weight_decay * tf.nn.l2_loss(v)
for v in tf.trainable_variables()
if 'weights' in v.name
]
# Loss function
all_loss.append(
tf.reduce_mean(loss) + tf.add_n(l2_losses))
tf.get_variable_scope().reuse_variables()
# Output size
#output_size = (self.raw_output.shape[1].value, self.raw_output.shape[2].value)
# Variables that load from pre-trained model.
# For training, last few layers should not be loaded.
if self.conf.pretrain_file is not None:
restore_var = [
v for v in tf.global_variables() if 'fc' not in v.name
]
original_step = int(self.conf.pretrain_file.split("-")[-1])
else:
original_step = 0
num_steps = self.conf.num_steps + original_step
# Trainable Variables
# Note that is_training=False still updates BN parameters gamma (scale) and beta (offset)
# if they are presented in var_list of the optimiser definition.
# So we remove them from the list.
all_trainable = [
v for v in tf.trainable_variables()
if 'beta' not in v.name and 'gamma' not in v.name
]
# Fine-tune part
conv_trainable = [v for v in all_trainable
if 'fc' not in v.name] # lr * 1.0
# ASPP part
fc_trainable = [v for v in all_trainable if 'fc' in v.name]
# fc_w_trainable = [v for v in fc_trainable if 'weights' in v.name] # lr * 10.0
# fc_b_trainable = [v for v in fc_trainable if 'biases' in v.name] # lr * 20.0
# check
#assert (len(all_trainable) == len(fc_trainable) + len(conv_trainable))
#assert (len(fc_trainable) == len(fc_w_trainable) + len(fc_b_trainable))
# Groud Truth: ignoring all labels greater or equal than n_classes
#label_proc = prepare_label(self.label_batch, output_size, num_classes=self.conf.num_classes,
#one_hot=False) # [batch_size, 41, 41]
#raw_gt = tf.reshape(label_proc, [-1, ])
#indices = tf.squeeze(tf.where(tf.less_equal(raw_gt, self.conf.num_classes - 1)), 1)
#gt = tf.cast(tf.gather(raw_gt, indices), tf.int32)
#raw_prediction = tf.reshape(self.raw_output, [-1, self.conf.num_classes])
#prediction = tf.gather(raw_prediction, indices)
# Pixel-wise softmax_cross_entropy loss
#loss = tf.nn.sparse_softmax_cross_entropy_with_logits(logits=prediction, labels=gt)
# L2 regularization
#l2_losses = [self.conf.weight_decay * tf.nn.l2_loss(v) for v in tf.trainable_variables() if 'weights' in v.name]
# Loss function
self.reduced_loss = tf.add_n(all_loss) / self.n_gpu
#self.reduced_loss = tf.reduce_mean(loss) + tf.add_n(l2_losses)
# Define optimizers
# 'poly' learning rate
base_lr = tf.constant(self.conf.learning_rate)
self.curr_step = tf.placeholder(dtype=tf.float32, shape=())
self.loss_trans = tf.placeholder(dtype=tf.float32, shape=())
self.final_loss = (self.reduced_loss + self.loss_trans) / 2
learning_rate = tf.scalar_mul(
base_lr, tf.pow((1 - self.curr_step / num_steps), self.conf.power))
#print self.conf.power
self.learning_rate = learning_rate
#print learning_rate
# We have several optimizers here in order to handle the different lr_mult
# which is a kind of parameters in Caffe. This controls the actual lr for each
# layer.
opt = tf.train.AdamOptimizer(learning_rate, self.conf.momentum, 0.98)
#opt= tf.train.MomentumOptimizer(learning_rate, self.conf.momentum)
#opt_fc_w = tf.train.AdamOptimizer(learning_rate , self.conf.momentum,0.98)
#opt_fc_b = tf.train.AdamOptimizer(learning_rate , self.conf.momentum,0.98)
#opt_conv = tf.train.MomentumOptimizer(learning_rate, self.conf.momentum)
#opt_fc_w = tf.train.MomentumOptimizer(learning_rate * 10.0, self.conf.momentum)
#opt_fc_b = tf.train.MomentumOptimizer(learning_rate * 20.0, self.conf.momentum)
# To make sure each layer gets updated by different lr's, we do not use 'minimize' here.
# Instead, we separate the steps compute_grads+update_params.
# Compute grads
grads_conv = tf.gradients(self.final_loss, conv_trainable)
# train_op = opt.apply_gradients(zip(grads_conv, conv_trainable))
#grads = tf.gradients(self.reduced_loss, conv_trainable + fc_w_trainable + fc_b_trainable)
grads_conv = grads_conv[:len(conv_trainable)]
# grads_fc_w = grads[len(conv_trainable): (len(conv_trainable) + len(fc_w_trainable))]
# grads_fc_b = grads[(len(conv_trainable) + len(fc_w_trainable)):]
# Update params
train_op_conv = opt.apply_gradients(zip(grads_conv, conv_trainable))
# train_op_conv = opt_conv.apply_gradients(zip(grads_conv, conv_trainable))
# train_op_fc_w = opt_fc_w.apply_gradients(zip(grads_fc_w, fc_w_trainable))
# train_op_fc_b = opt_fc_b.apply_gradients(zip(grads_fc_b, fc_b_trainable))
# Finally, get the train_op!
self.train_op = train_op_conv
# Saver for storing checkpoints of the model
self.saver = tf.train.Saver(var_list=tf.global_variables(),
max_to_keep=0)
# Loader for loading the pre-trained model
if self.conf.pretrain_file is not None:
self.loader = tf.train.Saver(var_list=restore_var)
def main():
"""Create the model and start the training."""
h = args.input_size
w = args.input_size
input_size = (h, w)
tf.set_random_seed(args.random_seed)
# Create queue coordinator.
coord = tf.train.Coordinator()
image_batch, label_batch = read_data(is_training=True)
# Create network.
net, end_points = deeplabv3(
image_batch,
num_classes=args.num_classes,
depth=args.num_layers,
is_training=True,
)
# For a small batch size, it is better to keep
# the statistics of the BN layers (running means and variances)
# frozen, and to not update the values provided by the pre-trained model.
# If is_training=True, the statistics will be updated during the training.
# Note that is_training=False still updates BN parameters gamma (scale)
# and beta (offset)
# if they are presented in var_list of the optimizer definition.
# Predictions.
raw_output = end_points['resnet{}/logits'.format(args.num_layers)]
# Which variables to load. Running means and variances are not trainable,
# thus all_variables() should be restored.
restore_var = [
v for v in tf.global_variables()
if 'fc' not in v.name or not args.not_restore_last
]
if args.freeze_bn:
all_trainable = [
v for v in tf.trainable_variables()
if 'beta' not in v.name and 'gamma' not in v.name
]
else:
all_trainable = [v for v in tf.trainable_variables()]
conv_trainable = [v for v in all_trainable if 'fc' not in v.name]
# Upsample the logits instead of donwsample the ground truth
raw_output_up = tf.image.resize_bilinear(raw_output, [h, w])
# Predictions: ignoring all predictions with labels greater or equal than
# n_classes
label_proc = tf.squeeze(label_batch)
mask = label_proc <= args.num_classes
seg_logits = tf.boolean_mask(raw_output_up, mask)
seg_gt = tf.boolean_mask(label_proc, mask)
seg_gt = tf.cast(seg_gt, tf.int32)
# Pixel-wise softmax loss.
loss = tf.nn.sparse_softmax_cross_entropy_with_logits(logits=seg_logits,
labels=seg_gt)
seg_loss = tf.reduce_mean(loss)
seg_loss_sum = tf.summary.scalar('loss/seg', seg_loss)
reg_losses = tf.get_collection(tf.GraphKeys.REGULARIZATION_LOSSES)
reg_loss = tf.add_n(reg_losses)
reg_loss_sum = tf.summary.scalar('loss/reg', reg_loss)
tot_loss = seg_loss + reg_loss
tot_loss_sum = tf.summary.scalar('loss/tot', tot_loss)
seg_pred = tf.argmax(seg_logits, axis=1)
train_mean_iou, train_update_mean_iou = streaming_mean_iou(
seg_pred, seg_gt, args.num_classes, name="train_iou")
train_iou_sum = tf.summary.scalar('accuracy/train_mean_iou',
train_mean_iou)
train_initializer = tf.variables_initializer(var_list=tf.get_collection(
tf.GraphKeys.LOCAL_VARIABLES, scope="train_iou"))
# Define loss and optimisation parameters.
base_lr = tf.constant(args.learning_rate)
step_ph = tf.placeholder(dtype=tf.float32, shape=())
learning_rate = tf.scalar_mul(
base_lr, tf.pow((1 - step_ph / args.num_steps), args.power))
# learning_rate = base_lr
lr_sum = tf.summary.scalar('params/learning_rate', learning_rate)
train_sum_op = tf.summary.merge(
[seg_loss_sum, reg_loss_sum, tot_loss_sum, train_iou_sum, lr_sum])
image_batch_val, label_batch_val = read_data(is_training=False)
_, end_points_val = deeplabv3(
image_batch_val,
num_classes=args.num_classes,
depth=args.num_layers,
reuse=True,
is_training=False,
)
raw_output_val = end_points_val['resnet{}/logits'.format(args.num_layers)]
nh, nw = tf.shape(image_batch_val)[1], tf.shape(image_batch_val)[2]
seg_logits_val = tf.image.resize_bilinear(raw_output_val, [nh, nw])
seg_pred_val = tf.argmax(seg_logits_val, axis=3)
seg_pred_val = tf.expand_dims(seg_pred_val, 3)
seg_pred_val = tf.reshape(seg_pred_val, [
-1,
])
seg_gt_val = tf.cast(label_batch_val, tf.int32)
seg_gt_val = tf.reshape(seg_gt_val, [
-1,
])
mask_val = seg_gt_val <= args.num_classes - 1
seg_pred_val = tf.boolean_mask(seg_pred_val, mask_val)
seg_gt_val = tf.boolean_mask(seg_gt_val, mask_val)
val_mean_iou, val_update_mean_iou = streaming_mean_iou(
seg_pred_val, seg_gt_val, num_classes=args.num_classes, name="val_iou")
val_iou_sum = tf.summary.scalar('accuracy/val_mean_iou', val_mean_iou)
val_initializer = tf.variables_initializer(var_list=tf.get_collection(
tf.GraphKeys.LOCAL_VARIABLES, scope="val_iou"))
test_sum_op = tf.summary.merge([val_iou_sum])
global_step = tf.train.get_or_create_global_step()
opt = tf.train.MomentumOptimizer(learning_rate, args.momentum)
grads_conv = tf.gradients(tot_loss, conv_trainable)
# train_op = opt.apply_gradients(zip(grads_conv, conv_trainable))
train_op = slim.learning.create_train_op(tot_loss,
opt,
global_step=global_step,
variables_to_train=conv_trainable,
summarize_gradients=True)
# Set up tf session and initialize variables.
config = tf.ConfigProto()
config.gpu_options.allow_growth = True
sess = tf.Session(config=config)
sess.run(tf.global_variables_initializer())
sess.run(tf.local_variables_initializer())
# Saver for storing checkpoints of the model.
saver = tf.train.Saver(var_list=tf.global_variables(), max_to_keep=20)
# Load variables if the checkpoint is provided.
if args.ckpt > 0 or args.restore_from is not None:
loader = tf.train.Saver(var_list=restore_var)
load(loader, sess, args.snapshot_dir)
# Start queue threads.
threads = tf.train.start_queue_runners(coord=coord, sess=sess)
# tf.get_default_graph().finalize()
summary_writer = tf.summary.FileWriter(args.snapshot_dir, sess.graph)
# Iterate over training steps.
for step in range(args.ckpt, args.num_steps):
start_time = time.time()
feed_dict = {step_ph: step}
tot_loss_float, seg_loss_float, reg_loss_float, _, lr_float, _, train_summary = sess.run(
[
tot_loss, seg_loss, reg_loss, train_op, learning_rate,
train_update_mean_iou, train_sum_op
],
feed_dict=feed_dict)
train_mean_iou_float = sess.run(train_mean_iou)
duration = time.time() - start_time
sys.stdout.write('step {:d}, tot_loss = {:.6f}, seg_loss = {:.6f}, ' \
'reg_loss = {:.6f}, mean_iou = {:.6f}, lr: {:.6f}({:.3f}' \
'sec/step)\n'.format(step, tot_loss_float, seg_loss_float,
reg_loss_float, train_mean_iou_float, lr_float, duration)
)
sys.stdout.flush()
if step % args.save_pred_every == 0 and step > args.ckpt:
summary_writer.add_summary(train_summary, step)
sess.run(val_initializer)
for val_step in range(NUM_VAL - 1):
_, test_summary = sess.run([val_update_mean_iou, test_sum_op],
feed_dict=feed_dict)
summary_writer.add_summary(test_summary, step)
val_mean_iou_float = sess.run(val_mean_iou)
save(saver, sess, args.snapshot_dir, step)
sys.stdout.write('step {:d}, train_mean_iou: {:.6f}, ' \
'val_mean_iou: {:.6f}\n'.format(step, train_mean_iou_float,
val_mean_iou_float))
sys.stdout.flush()
sess.run(train_initializer)
if coord.should_stop():
coord.request_stop()
coord.join(threads)
def main(snapshot_dir, image_path, output_path):
"""Create the model and start the training."""
tf.set_random_seed(args.random_seed)
# Create queue coordinator.
coord = tf.train.Coordinator()
os.environ['CUDA_VISIBLE_DEVICES'] = '2'
image = tf.placeholder(dtype=tf.int32, shape=[None, None, 3], name='image')
image_float = tf.to_float(image)
image_reverse = tf.reverse(image_float, axis=[-1])
image_norm = image_reverse - IMG_MEAN
image_batch = tf.expand_dims(image_norm, axis=0)
# Create network.
net, end_points = deeplabv3(image_batch,
num_classes=args.num_classes,
depth=args.num_layers,
is_training=False)
restore_var = [
v for v in tf.global_variables()
if 'fc' not in v.name or not args.not_restore_last
]
# Predictions.
raw_output = end_points['resnet_v1_{}/logits'.format(args.num_layers)]
out_put_score = tf.nn.softmax(raw_output, axis=3)
out_put_score = tf.reduce_max(out_put_score, axis=3)
seg_pred = tf.argmax(raw_output, axis=3, output_type=tf.int32)
# Set up tf session and initialize variables.
config = tf.ConfigProto()
config.gpu_options.allow_growth = True
sess = tf.Session(config=config)
sess.run(tf.global_variables_initializer())
sess.run(tf.local_variables_initializer())
# Load variables if the checkpoint is provided.
# if args.ckpt > 0 or args.restore_from is not None:
loader = tf.train.Saver(var_list=restore_var)
load(loader, sess, snapshot_dir)
graph = tf.get_default_graph()
input_graph_def = graph.as_graph_def()
# Start queue threads.
threads = tf.train.start_queue_runners(coord=coord, sess=sess)
tf.get_default_graph().finalize()
file_list = os.listdir(image_path)
# image_file = file_list[0]
image_file = '/home/gq/xnetsrc/image/id00018.jpg'
file_path = os.path.join(image_path, image_file)
image_data = cv2.imread(file_path, cv2.IMREAD_COLOR)
_, seg = sess.run([out_put_score, seg_pred], {image: image_data})
output_node_names = ['Max', 'ArgMax']
output_graph_def = graph_util.convert_variables_to_constants(
sess, input_graph_def, output_node_names)
pb_file = os.path.join(snapshot_dir, "model.pb")
with tf.gfile.GFile(pb_file, "wb") as f:
f.write(output_graph_def.SerializeToString())
print("%d ops in the final graph." % len(output_graph_def.node))
with tf.gfile.GFile(pb_file, "rb") as f:
graph_def = tf.GraphDef()
graph_def.ParseFromString(f.read())
def print_ops():
with tf.Graph().as_default() as graph:
tf.import_graph_def(graph_def,
input_map=None,
return_elements=None,
name="prefix",
op_dict=None,
producer_op_list=None)
for op in graph.get_operations():
print(op.name, op.values())
print_ops()
if coord.should_stop():
coord.request_stop()
coord.join(threads)
def main():
"""Create the model and start the training."""
h = args.input_size
w = args.input_size
input_size = (h, w)
tf.set_random_seed(args.random_seed)
# Create queue coordinator.
coord = tf.train.Coordinator()
image_batch, label_batch = read_data()
# Create network.
net, end_points = deeplabv3(image_batch,
num_classes=args.num_classes,
layer_depth=args.num_layers,
is_training=args.is_training)
# For a small batch size, it is better to keep
# the statistics of the BN layers (running means and variances)
# frozen, and to not update the values provided by the pre-trained model.
# If is_training=True, the statistics will be updated during the training.
# Note that is_training=False still updates BN parameters gamma (scale) and beta (offset)
# if they are presented in var_list of the optimiser definition.
# Predictions.
raw_output = end_points['resnet{}/logits'.format(args.num_layers)]
# Which variables to load. Running means and variances are not trainable,
# thus all_variables() should be restored.
restore_var = [
v for v in tf.global_variables()
if 'fc' not in v.name or not args.not_restore_last
]
all_trainable = [
v for v in tf.trainable_variables()
if 'beta' not in v.name and 'gamma' not in v.name
]
conv_trainable = [v for v in all_trainable
if 'fc' not in v.name] # lr * 1.0
# Predictions: ignoring all predictions with labels greater or equal than n_classes
label_proc = prepare_label(label_batch,
tf.shape(raw_output)[1:3],
args.num_classes,
one_hot=False)
mask = label_proc <= args.num_classes
seg_logits = tf.boolean_mask(raw_output, mask)
seg_gt = tf.boolean_mask(label_proc, mask)
seg_gt = tf.cast(seg_gt, tf.int32)
# Pixel-wise softmax loss.
loss = tf.nn.sparse_softmax_cross_entropy_with_logits(logits=seg_logits,
labels=seg_gt)
seg_loss = tf.reduce_mean(loss)
tf.summary.scalar('loss/seg', seg_loss)
reg_losses = tf.get_collection(tf.GraphKeys.REGULARIZATION_LOSSES)
reg_loss = args.weight_decay * tf.add_n(reg_losses)
tf.summary.scalar('loss/reg', reg_loss)
total_loss = seg_loss + reg_loss
tf.summary.scalar('loss/tot', total_loss)
seg_pred = tf.argmax(seg_logits, axis=1)
mean_iou, update_mean_iou = streaming_mean_iou(seg_pred, seg_gt,
args.num_classes)
tf.summary.scalar('accuracy/mean_iou', mean_iou)
summary_op = tf.summary.merge_all()
global_step = slim.get_or_create_global_step()
# Define loss and optimisation parameters.
base_lr = tf.constant(args.learning_rate)
step_ph = tf.placeholder(dtype=tf.float32, shape=())
learning_rate = tf.scalar_mul(
base_lr, tf.pow((1 - step_ph / args.num_steps), args.power))
# learning_rate = base_lr
tf.summary.scalar('hyperparameters/learning_rate', learning_rate)
opt = tf.train.MomentumOptimizer(learning_rate, args.momentum)
grads_conv = tf.gradients(total_loss, conv_trainable)
# train_op = opt.apply_gradients(zip(grads_conv, conv_trainable))
train_op = slim.learning.create_train_op(total_loss,
opt,
global_step=global_step,
variables_to_train=conv_trainable,
summarize_gradients=True)
# Set up tf session and initialize variables.
config = tf.ConfigProto()
config.gpu_options.allow_growth = True
sess = tf.Session(config=config)
sess.run(tf.global_variables_initializer())
sess.run(tf.local_variables_initializer())
# Saver for storing checkpoints of the model.
saver = tf.train.Saver(var_list=tf.global_variables(), max_to_keep=20)
# Load variables if the checkpoint is provided.
if args.ckpt > 0 or args.restore_from is not None:
loader = tf.train.Saver(var_list=restore_var)
load(loader, sess, args.snapshot_dir)
# Start queue threads.
threads = tf.train.start_queue_runners(coord=coord, sess=sess)
tf.get_default_graph().finalize()
summary_writer = tf.summary.FileWriter(args.snapshot_dir, sess.graph)
# Iterate over training steps.
for step in range(args.ckpt, args.num_steps):
# for step in range(1):
start_time = time.time()
feed_dict = {step_ph: step}
if step % args.save_pred_every == 0 and step > args.ckpt:
tot_loss_float, seg_loss_float, reg_loss_float, images, labels, summary, mean_iou_float, _, _, lr_float = sess.run(
[
total_loss, seg_loss, reg_loss, image_batch, label_batch,
summary_op, mean_iou, update_mean_iou, train_op,
learning_rate
],
feed_dict=feed_dict)
summary_writer.add_summary(summary, step)
save(saver, sess, args.snapshot_dir, step)
sys.stdout.write(
'step {:d}, tot_loss = {:.6f}, seg_loss = {:.6f}, reg_loss = {:.6f}, mean_iou: {:.6f}, lr: {:.6f}({:.3f} sec/step)\n'
.format(step, tot_loss_float, seg_loss_float, reg_loss_float,
mean_iou_float, lr_float, duration))
sys.stdout.flush()
else:
tot_loss_float, seg_loss_float, reg_loss_float, mean_iou_float, _, _, lr_float = sess.run(
[
total_loss, seg_loss, reg_loss, mean_iou, update_mean_iou,
train_op, learning_rate
],
feed_dict=feed_dict)
duration = time.time() - start_time
sys.stdout.write(
'step {:d}, tot_loss = {:.6f}, seg_loss = {:.6f}, reg_loss = {:.6f}, mean_iou: {:.6f}, lr: {:.6f}({:.3f} sec/step)\n'
.format(step, tot_loss_float, seg_loss_float, reg_loss_float,
mean_iou_float, lr_float, duration))
sys.stdout.flush()
if coord.should_stop():
coord.request_stop()
coord.join(threads)
def main():
"""Create the model and start the training."""
# h = args.input_size
# w = args.input_size
#input_size = (h, w)
tf.set_random_seed(args.random_seed)
# Create queue coordinator.
coord = tf.train.Coordinator()
image_batch, label_batch = read_data(is_training=args.is_training,
split_name=args.split_name)
cropped_image = tf.summary.image('cropped_image', image_batch)
coppped_label = tf.summary.image('cropped_label', label_batch * 255)
# Create network.
net, end_points = deeplabv3(
image_batch,
num_classes=args.num_classes,
depth=args.num_layers,
is_training=args.is_training,
)
# For a small batch size, it is better to keep
# the statistics of the BN layers (running means and variances)
# frozen, and to not update the values provided by the pre-trained model.
# If is_training=True, the statistics will be updated during the training.
# Note that is_training=False still updates BN parameters gamma (scale)
# and beta (offset)
# if they are presented in var_list of the optimizer definition.
# Predictions.
raw_output = end_points['resnet_v1_{}/logits'.format(args.num_layers)]
# Which variables to load. Running means and variances are not trainable,
# thus all_variables() should be restored.
if args.imagenet is not None and args.ckpt == 0:
restore_var = [
v for v in tf.global_variables()
if ('aspp' not in v.name) and ('img_pool' not in v.name) and (
'fusion' not in v.name) and ('block5' not in v.name) and (
'block6' not in v.name) and ('block7' not in v.name) and (
'logits' not in v.name) and ('short_cut' not in v.name)
]
else:
restore_var = [v for v in tf.global_variables()]
if args.freeze_bn:
all_trainable = [
v for v in tf.trainable_variables()
if 'beta' not in v.name and 'gamma' not in v.name
]
else:
all_trainable = [v for v in tf.trainable_variables()]
conv_trainable = [v for v in all_trainable]
# Upsample the logits instead of donwsample the ground truth
# raw_output_up = tf.image.resize_bilinear(raw_output, [h, w])
# Predictions: ignoring all predictions with labels greater or equal than
# n_classes
label_proc = tf.squeeze(label_batch)
mask = label_proc <= args.num_classes
seg_logits = tf.boolean_mask(raw_output, mask)
seg_gt = tf.boolean_mask(label_proc, mask)
seg_gt = tf.cast(seg_gt, tf.int32)
# Pixel-wise softmax loss.
loss = tf.nn.sparse_softmax_cross_entropy_with_logits(logits=seg_logits,
labels=seg_gt)
seg_loss = tf.reduce_mean(loss)
seg_loss_sum = tf.summary.scalar('loss/seg', seg_loss)
reg_losses = tf.get_collection(tf.GraphKeys.REGULARIZATION_LOSSES)
reg_loss = tf.add_n(reg_losses)
reg_loss_sum = tf.summary.scalar('loss/reg', reg_loss)
tot_loss = seg_loss + 2 * reg_loss
tot_loss_sum = tf.summary.scalar('loss/tot', tot_loss)
# 通过argmax来获取预测结果
seg_pred = tf.argmax(seg_logits, axis=1)
train_mean_iou, train_update_mean_iou = streaming_mean_iou(
seg_pred, seg_gt, args.num_classes, name="train_iou")
train_iou_sum = tf.summary.scalar('accuracy/train_mean_iou',
train_mean_iou)
train_initializer = tf.variables_initializer(var_list=tf.get_collection(
tf.GraphKeys.LOCAL_VARIABLES, scope="train_iou"))
# Define loss and optimisation parameters.
base_lr = tf.constant(args.learning_rate)
step_ph = tf.placeholder(dtype=tf.float32, shape=())
learning_rate = tf.scalar_mul(
base_lr, tf.pow((1 - step_ph / args.num_steps), args.power))
# learning_rate = base_lr
lr_sum = tf.summary.scalar('params/learning_rate', learning_rate)
train_sum_op = tf.summary.merge([
seg_loss_sum, reg_loss_sum, tot_loss_sum, train_iou_sum, lr_sum,
cropped_image, coppped_label
])
image_batch_val, label_batch_val = read_data(is_training=False,
split_name='val')
_, end_points_val = deeplabv3(
image_batch_val,
num_classes=args.num_classes,
depth=args.num_layers,
reuse=True,
is_training=False,
)
raw_output_val = end_points_val['resnet_v1_{}/logits'.format(
args.num_layers)]
nh, nw = tf.shape(image_batch_val)[1], tf.shape(image_batch_val)[2]
seg_logits_val = tf.image.resize_bilinear(raw_output_val, [nh, nw])
seg_pred_val = tf.argmax(seg_logits_val, axis=3)
seg_pred_val = tf.expand_dims(seg_pred_val, 3)
seg_pred_val = tf.reshape(seg_pred_val, [
-1,
])
seg_gt_val = tf.cast(label_batch_val, tf.int32)
seg_gt_val = tf.reshape(seg_gt_val, [
-1,
])
mask_val = seg_gt_val <= args.num_classes - 1
seg_pred_val = tf.boolean_mask(seg_pred_val, mask_val)
seg_gt_val = tf.boolean_mask(seg_gt_val, mask_val)
val_mean_iou, val_update_mean_iou = streaming_mean_iou(
seg_pred_val, seg_gt_val, num_classes=args.num_classes, name="val_iou")
val_iou_sum = tf.summary.scalar('accuracy/val_mean_iou', val_mean_iou)
val_initializer = tf.variables_initializer(var_list=tf.get_collection(
tf.GraphKeys.LOCAL_VARIABLES, scope="val_iou"))
test_sum_op = tf.summary.merge([val_iou_sum])
global_step = tf.train.get_or_create_global_step()
opt = tf.train.MomentumOptimizer(learning_rate, args.momentum)
# grads_conv = tf.gradients(tot_loss, conv_trainable)
# train_op = opt.apply_gradients(zip(grads_conv, conv_trainable))
train_op = slim.learning.create_train_op(tot_loss,
opt,
global_step=global_step,
variables_to_train=conv_trainable,
summarize_gradients=True)
# Set up tf session and initialize variables.
config = tf.ConfigProto()
config.gpu_options.allow_growth = True
sess = tf.Session(config=config)
sess.run(tf.global_variables_initializer())
sess.run(tf.local_variables_initializer())
# Saver for storing checkpoints of the model.
saver = tf.train.Saver(var_list=tf.global_variables(), max_to_keep=50)
# Load variables if the checkpoint is provided.
if args.ckpt > 0 or args.restore_from is not None or args.imagenet is not None:
loader = tf.train.Saver(var_list=restore_var)
load(loader, sess, args.snapshot_dir)
# Start queue threads.
threads = tf.train.start_queue_runners(coord=coord, sess=sess)
# tf.get_default_graph().finalize()
summary_writer = tf.summary.FileWriter(args.snapshot_dir, sess.graph)
# Iterate over training steps.
for step in range(args.ckpt, args.num_steps + 1):
start_time = time.time()
feed_dict = {step_ph: step}
tot_loss_float, seg_loss_float, reg_loss_float, _, lr_float, _, train_summary = sess.run(
[
tot_loss, seg_loss, reg_loss, train_op, learning_rate,
train_update_mean_iou, train_sum_op
],
feed_dict=feed_dict)
train_mean_iou_float = sess.run(train_mean_iou)
summary_writer.add_summary(train_summary, step)
# batch_prediction = sess.run(raw_output_up)
# batch_image = sess.run(image_batch)
# with open('./datasets/IDImage/{}.txt'.format(args.split_name), 'r') as f:
# lines = f.readlines()
# for i in range(args.batch_size):
# pre_tosave = batch_prediction[i,:,:,:]
# pre_tosave = np.argmax(pre_tosave, axis=2)
# img_tosave = batch_image[i,:,:,:]
# img_tosave = np.squeeze(img_tosave)
# img_tosave = img_tosave[:,:,[2,1,0]]
#
#
# save_name = lines[(step-args.ckpt)*args.batch_size + i].split('/')[-1][:-1]
#
# img_save_path = os.path.join('./datasets/IDImage/test/JPGImage',save_name)
# pre_save_path = os.path.join('./datasets/IDImage/test/Prediction',save_name)
# pyplot.imsave(img_save_path, img_tosave)
# pyplot.imsave(pre_save_path,pre_tosave)
duration = time.time() - start_time
sys.stdout.write('step {:d}, tot_loss = {:.6f}, seg_loss = {:.6f}, ' \
'reg_loss = {:.6f}, mean_iou = {:.6f}, lr: {:.6f}({:.3f}' \
'sec/step), memory usage: {}\n'.format(step, tot_loss_float, seg_loss_float,
reg_loss_float, train_mean_iou_float, lr_float, duration,memory_usage_psutil())
)
sys.stdout.flush()
if step % args.save_pred_every == 0 and step > args.ckpt:
sess.run(val_initializer)
val_mean_iou_total = 0
for val_step in range(NUM_VAL):
_, test_summary, val_mean_iou_single = sess.run(
[val_update_mean_iou, test_sum_op, val_mean_iou],
feed_dict=feed_dict)
val_mean_iou_total += val_mean_iou_single
test_mean_iou = val_mean_iou_total / (NUM_VAL - 1)
print('=============================================')
print(test_mean_iou)
print('=============================================')
if not os.path.exists('./datasets/IDImage/logs.txt'):
os.system("touch ./datasets/IDImage/logs.txt")
with open('./datasets/IDImage/logs.txt', 'a+') as f:
f.write('{}steps: {}'.format(step, test_mean_iou) + '\n')
# val_mean_iou_mean = tf.summary.scalar('val/mean_iou', val_mean_iou_float)
# val_sum_op = tf.summary.merge([val_mean_iou_mean])
# val_summary = sess.run(val_sum_op)
# summary_writer.add_summary(val_summary, step)
summary_writer.add_summary(test_summary, step)
val_mean_iou_float = sess.run(val_mean_iou)
save(saver, sess, args.snapshot_dir, step)
sys.stdout.write('step {:d}, train_mean_iou: {:.6f}, ' \
'val_mean_iou: {:.6f}\n'.format(step, train_mean_iou_float,
val_mean_iou_float))
sys.stdout.flush()
sess.run(train_initializer)
if coord.should_stop():
coord.request_stop()
coord.join(threads)
