num_classes = 249
dataset_name = 'isogr_ rgb'
training_datalist = 'trte_splits/IsoGD_Image/train_rgb_list.txt'
testing_datalist = 'trte_splits/IsoGD_Image/valid_rgb_list.txt'
sess = tf.InteractiveSession()
#x=[batch_size,seq_len,height,width,channels]
x = tf.placeholder(tf.float32, [batch_size, seq_len, 112, 112, 3],
name='datas')
# y=[batch_size,label]
y = tf.placeholder(tf.int32, shape=[
batch_size,
], name='labels')
# get the output of the layer
networks = net.c3d_clstm(x, num_classes, False, True)
networks_y = networks.outputs
print(networks)
networks_y_op = tf.argmax(tf.nn.softmax(networks_y), 1)
networks_cost = tl.cost.cross_entropy(networks_y, y, name="cost_network")
tf.summary.scalar("network loss", networks_cost)
correct_pred = tf.equal(tf.cast(networks_y_op, tf.int32), y)
networks_accu = tf.reduce_mean(tf.cast(correct_pred, tf.float32))
tf.summary.scalar("accary", networks_accu)
# test the model files
# predictions = net.c3d_clstm(x,num_classes,True,False)
# predictions_y_op = tf.argmax(tf.nn.softmax(predictions.outputs),1)
# predictions_accu = tf.reduce_mean(tf.cast(tf.equal(tf.cast(predictions_y_op),tf.int32),y),tf.float32)
strtime = '%s%s%s-%s%s%s' % (d.split('-')[0], d.split('-')[1], d.split('-')[2],
t.split(':')[0], t.split(':')[1], t.split(':')[2])
saved_stdout = sys.stdout
mem_log = cStringIO.StringIO()
sys.stdout = mem_log
logfile = './log/training_%s_%s.log' % (dataset_name, strtime)
log = open(logfile, 'w')
sess = tf.InteractiveSession()
x = tf.placeholder(tf.float32, [batch_size, seq_len, 112, 112, 3], name='x')
y = tf.placeholder(tf.int32, shape=[
batch_size,
], name='y')
networks = net.c3d_clstm(x, num_classes, False, True)
networks_y = networks.outputs
networks_y_op = tf.argmax(tf.nn.softmax(networks_y), 1)
networks_cost = tl.cost.cross_entropy(networks_y, y, 'loss')
correct_pred = tf.equal(tf.cast(networks_y_op, tf.int32), y)
networks_accu = tf.reduce_mean(tf.cast(correct_pred, tf.float32))
predictons = net.c3d_clstm(x, num_classes, True, False)
predicton_y_op = tf.argmax(tf.nn.softmax(predictons.outputs), 1)
predicton_accu = tf.reduce_mean(
tf.cast(tf.equal(tf.cast(predicton_y_op, tf.int32), y), tf.float32))
l2_cost = tf.contrib.layers.l2_regularizer(weight_decay)(networks.all_params[0]) + \
tf.contrib.layers.l2_regularizer(weight_decay)(networks.all_params[6]) + \
tf.contrib.layers.l2_regularizer(weight_decay)(networks.all_params[12]) + \
tf.contrib.layers.l2_regularizer(weight_decay)(networks.all_params[14]) + \
