Which way is ‘right’?: Uncovering limitations of Vision-and-Language Navigation models

Related