We study the automatic generation of navigation instructions from 360-degree images captured on indoor routes. Existing generators suffer from poor visual grounding, causing them to rely on language priors and hallucinate objects. Our MARKY-MT5 system addresses this by focusing on visual landmarks; it comprises a first stage landmark detector and a second stage generator -- a multimodal, multilingual, multitask encoder-decoder. To train it, we bootstrap grounded landmark annotations on top of the Room-across-Room (RxR) dataset. Using text parsers, weak supervision from RxR's pose traces, and a multilingual image-text encoder trained on 1.8b images, we identify 971k English, Hindi and Telugu landmark descriptions and ground them to specific regions in panoramas. On Room-to-Room, human wayfinders obtain success rates (SR) of 71% following MARKY-MT5's instructions, just shy of their 75% SR following human instructions -- and well above SRs with other generators. Evaluations on RxR's longer, diverse paths obtain 61-64% SRs on three languages. Generating such high-quality navigation instructions in novel environments is a step towards conversational navigation tools and could facilitate larger-scale training of instruction-following agents.
@article{arxiv.2111.12872,
title = {Less is More: Generating Grounded Navigation Instructions from Landmarks},
author = {Su Wang and Ceslee Montgomery and Jordi Orbay and Vighnesh Birodkar and Aleksandra Faust and Izzeddin Gur and Natasha Jaques and Austin Waters and Jason Baldridge and Peter Anderson},
journal= {arXiv preprint arXiv:2111.12872},
year = {2022}
}