1 paper
Zijiao Yang, Arjun Majumdar, Stefan Lee
To be successful, Vision-and-Language Navigation (VLN) agents must be able to ground instructions to actions based on their surroundings. In this work, we develop a methodology to…