Possible bug in using Batch normalisation

[mrtpk]

• training parameter is not given while initialising tf.layers.batch_normalization. The training parameter defaults to False. Hence while training the moving mean and moving variance won't be updated.

• Since this branch's tf api is <tf2, tf.GraphKeys.UPDATE_OPS and train_op have to be executed before training [ref]. I could not find these updations in the training code.

CRNN/CRNN/crnn.py

Line 163 in becd222

bnorm1 = tf.layers.batch_normalization(conv3)

[lulupanPPY]

boy，you saved this code