Improve exporting of pre-trained models

[davidsandberg]

Importing of a pre-trained models using import_meta_graph(...) has been implemented here.
However, this code seems to initialize the graph with the instantaneous parameter values and not the EMA filtered ones, and this degrades LFW performance a bit.
Figure out how to import a graph and initialize it with filtered parameter values.

[davidsandberg]

This has now been fixed in validate_on_lfw.py.