Exploitation-Guided Exploration for Semantic Embodied Navigation
This work addresses the challenge of efficient and accurate goal-conditioned navigation for robots, representing an incremental improvement in modular policy frameworks.
The paper tackles the problem of semantic embodied navigation by proposing Exploitation-Guided Exploration (XGX), a method that combines exploration and exploitation modules to improve performance, increasing state-of-the-art accuracy from 70% to 73% on object navigation and showing over two-fold better sim-to-real transfer than baselines.
In the recent progress in embodied navigation and sim-to-robot transfer, modular policies have emerged as a de facto framework. However, there is more to compositionality beyond the decomposition of the learning load into modular components. In this work, we investigate a principled way to syntactically combine these components. Particularly, we propose Exploitation-Guided Exploration (XGX) where separate modules for exploration and exploitation come together in a novel and intuitive manner. We configure the exploitation module to take over in the deterministic final steps of navigation i.e. when the goal becomes visible. Crucially, an exploitation module teacher-forces the exploration module and continues driving an overridden policy optimization. XGX, with effective decomposition and novel guidance, improves the state-of-the-art performance on the challenging object navigation task from 70% to 73%. Along with better accuracy, through targeted analysis, we show that XGX is also more efficient at goal-conditioned exploration. Finally, we show sim-to-real transfer to robot hardware and XGX performs over two-fold better than the best baseline from simulation benchmarking. Project page: xgxvisnav.github.io