xiaoyang-coder
xiaoyang-coder
` def _bytes_feature(self, value): return tf.train.Feature(bytes_list=tf.train.BytesList(value=[value])) feature[key] = self._bytes_feature(np.array(data, dtype=np.float32).tobytes()) example_proto = tf.train.Example(features=tf.train.Features(feature=feature)) tf_writer.write(example_proto.SerializeToString()) ` Thank you very much.
I followed the instructions completely and used the data provided in Google drive, but the accuracy of ST-GCN was extremely low in NTU-RGB-D-xsub
Could you please implement the 3d OctConv