English

Neural Modular Control for Embodied Question Answering

Artificial Intelligence 2019-05-06 v2 Computation and Language Computer Vision and Pattern Recognition Machine Learning

Abstract

We present a modular approach for learning policies for navigation over long planning horizons from language input. Our hierarchical policy operates at multiple timescales, where the higher-level master policy proposes subgoals to be executed by specialized sub-policies. Our choice of subgoals is compositional and semantic, i.e. they can be sequentially combined in arbitrary orderings, and assume human-interpretable descriptions (e.g. 'exit room', 'find kitchen', 'find refrigerator', etc.). We use imitation learning to warm-start policies at each level of the hierarchy, dramatically increasing sample efficiency, followed by reinforcement learning. Independent reinforcement learning at each level of hierarchy enables sub-policies to adapt to consequences of their actions and recover from errors. Subsequent joint hierarchical training enables the master policy to adapt to the sub-policies. On the challenging EQA (Das et al., 2018) benchmark in House3D (Wu et al., 2018), requiring navigating diverse realistic indoor environments, our approach outperforms prior work by a significant margin, both in terms of navigation and question answering.

Keywords

Cite

@article{arxiv.1810.11181,
  title  = {Neural Modular Control for Embodied Question Answering},
  author = {Abhishek Das and Georgia Gkioxari and Stefan Lee and Devi Parikh and Dhruv Batra},
  journal= {arXiv preprint arXiv:1810.11181},
  year   = {2019}
}

Comments

10 pages, 3 figures, 2 tables. Published at CoRL 2018. Webpage: https://embodiedqa.org/

R2 v1 2026-06-23T04:53:20.154Z