Performance issues in /ScanNet (by P3)

[DLPerf]

Hello! I've found a performance issue in your project: batch() should be called before map(), which could make your program more efficient. Here is the tensorflow document to support it.

Detailed description is listed below:

• /train_scannet_grid.py: train_data.batch(FLAGS.batch_size, drop_remainder=True)(here) should be called before train_data.map(map_func=map_func, num_parallel_calls=dataset.num_threads)(here).
• /train_scannet_grid.py: val_data.batch(FLAGS.batch_size, drop_remainder=True)(here) should be called before val_data.map(map_func=map_func, num_parallel_calls=dataset.num_threads)(here).
• /test_scannet_grid.py: val_data.batch(FLAGS.batch_size, drop_remainder=True)(here) should be called before val_data.map(map_func=map_func, num_parallel_calls=dataset.num_threads)(here).

Besides, you need to check the function called in map()(e.g., map_func called in val_data.map(map_func=map_func, num_parallel_calls=dataset.num_threads)) whether to be affected or not to make the changed code work properly. For example, if map_func needs data with shape (x, y, z) as its input before fix, it would require data with shape (batch_size, x, y, z).

Looking forward to your reply. Btw, I am very glad to create a PR to fix it if you are too busy.

[yanx27]

Please feel free for creating a PR :)