Closed alvenchen closed 5 years ago
accourding to cnn_lstm_otc_ocr.py : self.seq_len = tf.fill([x.get_shape().as_list()[0]], feature_w)
that is seq_len equals batch_size. but why?
accourding to cnn_lstm_otc_ocr.py : self.seq_len = tf.fill([x.get_shape().as_list()[0]], feature_w)
that is seq_len equals batch_size. but why?