1 paper
Zehao Wang, Minye Wu, Yixin Cao +3
This study presents a novel evaluation framework for the Vision-Language Navigation (VLN) task. It aims to diagnose current models for various instruction categories at a finer-gra…