def model(self):
outdims = self.__self_dict()
fn = layers.l2_regularizer(1e-5)
fn0 = tf.no_regularizer
with slim.arg_scope([slim.conv2d, slim.fully_connected],
weights_regularizer=fn,
biases_regularizer=fn0,
normalizer_fn=slim.batch_norm):
with slim.arg_scope([slim.batch_norm],
is_training=False,
updates_collections=None,
decay=0.9,
center=True,
scale=True,
epsilon=1e-5):
pred_comb_ht, pred_comb_hand, pred_hand, pred_ht = basenet2(
self.inputs, kp=1, is_training=False, outdims=outdims)
self.hand_tensor = pred_hand
import tensorflow.contrib.layers as layers
fn = layers.l2_regularizer(1e-5)
fn0 = tf.no_regularizer
with slim.arg_scope([slim.conv2d, slim.fully_connected],
weights_regularizer=fn,
biases_regularizer=fn0,
normalizer_fn=slim.batch_norm):
with slim.arg_scope([slim.batch_norm],
is_training=is_train,
updates_collections=None,
decay=0.9,
center=True,
scale=True,
epsilon=1e-5):
pred_comb_ht, pred_comb_hand, pred_hand, pred_ht = basenet2(
inputs, kp=kp, is_training=is_train, outdims=outdims)
gt_palm_ht = tf.concat((gt_ht[:, :, :, 0:1], gt_ht[:, :, :, 1::4]), 3)
gt_fing_ht = tf.concat(
(gt_ht[:, :, :, 2::4], gt_ht[:, :, :, 3::4], gt_ht[:, :, :, 4::4]), 3)
label1 = tf.reshape(label, (-1, 21, 3))
gt_fing = tf.reshape(
tf.concat((label1[:, 2::4, :], label1[:, 3::4, :], label1[:, 4::4, :]), 1),
(-1, 15 * 3))
gt_palm = tf.reshape(tf.concat((label1[:, 0:1, :], label1[:, 1::4, :]), 1),
(-1, 6 * 3))
loss_ht = tf.nn.l2_loss((pred_ht - gt_ht)) / batch_size
loss_hand = tf.nn.l2_loss((pred_hand - label)) / batch_size
fn = layers.l2_regularizer(1e-5)
fn0 = tf.no_regularizer
with slim.arg_scope([slim.conv2d, slim.fully_connected],
weights_regularizer=fn,
biases_regularizer=fn0,
normalizer_fn=slim.batch_norm):
with slim.arg_scope([slim.batch_norm],
is_training=False,
updates_collections=None,
decay=0.9,
center=True,
scale=True,
epsilon=1e-5):
pred_comb_ht, pred_comb_hand, pred_hand, pred_ht = basenet2(
inputs, kp=1, is_training=False)
pred_out = pred_hand
import time
pred_norm = []
saver = tf.train.Saver(max_to_keep=5)
with tf.Session() as sess:
init = tf.global_variables_initializer()
sess.run(init)
saver.restore(sess, '../../model/crossInfoNet_NYU.ckpt')
loopv = test_num // batch_size
other = test_data[loopv * batch_size:]
a = time.time()
for i in xrange(loopv + 1):