Improving the Naturalness and Diversity of Referring Expression Generation models using Minimum Risk Training.

Nikolaos Panagiaris, Emma Hart, Dimitra Gkatzia

Browse the full INLG paper archive.