'is_training': is_training,
# Decay for the moving averages.
'decay': 0.995,
# epsilon to prevent 0s in variance.
'epsilon': 0.00001
}
with slim.arg_scope(
[slim.conv2d],
weights_initializer=slim.xavier_initializer_conv2d(
uniform=True),
weights_regularizer=slim.l2_regularizer(weight_decay),
normalizer_fn=slim.batch_norm,
normalizer_params=batch_norm_params):
with slim.arg_scope([slim.dropout], is_training=is_training) as sc:
return sc
if __name__ == '__main__':
import sys
sys.path.insert(0, '../../libs')
from tf_utils import print_endpoints
inputs = tf.placeholder(tf.float32, [1, 32, 256, 1], name="inputs")
is_training = tf.placeholder(tf.bool, name="is_training")
img_file = '/home/cwq/data/ocr/train_data/400w_eng_corpus/val/00000000.jpg'
squeeze_net = SqueezeNet(inputs, is_training)
print_endpoints(squeeze_net, inputs, is_training, img_file)
'is_training': is_training,
# Decay for the moving averages.
'decay': 0.995,
# epsilon to prevent 0s in variance.
'epsilon': 0.00001
}
with slim.arg_scope([slim.conv2d],
activation_fn=tf.nn.leaky_relu,
weights_regularizer=slim.l2_regularizer(0.0001),
weights_initializer=slim.xavier_initializer_conv2d(
uniform=True),
normalizer_fn=slim.batch_norm,
normalizer_params=batch_norm_params):
with slim.arg_scope([slim.dropout], is_training=is_training) as sc:
return sc
if __name__ == '__main__':
import sys
sys.path.insert(0, '../../libs')
from tf_utils import print_endpoints
inputs = tf.placeholder(tf.float32, [1, 32, 256, 1], name="inputs")
is_training = tf.placeholder(tf.bool, name="is_training")
img_file = '/home/cwq/data/ocr/train_data/400w_eng_corpus/val/00000000.jpg'
net = SimpleNet(inputs, is_training)
print_endpoints(net, inputs, is_training, img_file)
resnet_v1_block('block1', base_depth=64, num_units=3, stride=2),
resnet_v1_block('block2', base_depth=128, num_units=4, stride=2),
resnet_v1_block('block3', base_depth=256, num_units=6, stride=2),
resnet_v1_block('block4', base_depth=512, num_units=3, stride=1)
]
net, end_points = resnet_v1(inputs,
blocks,
is_training=is_training,
global_pool=False,
include_root_block=True,
scope=self._scope)
return net, end_points
if __name__ == '__main__':
import sys
sys.path.insert(0, '../libs')
from tf_utils import print_endpoints
img_file = '/home/cwq/data/VOCdevkit2007/VOC2007/JPEGImages/icdar13_100.jpg'
# img_file = '/home/cwq/ssd_data/more_bg_corpus/val/00000000.jpg'
res_net = ResNetV1()
res_net.create_architecture()
print_endpoints(res_net, img_file)
for k, v in res_net.end_points.items():
stride = math.ceil(800 / v.shape.as_list()[2])
print("%s, stride %d, shape %s" % (k, stride, v.shape.as_list()))