A Speech-First Model for Repair Detection and Correction

Christine Nakatani, Julia Hirschberg