We propose a simple but effective modular approach MOPA (Modular ObjectNav
with PointGoal agents) to systematically investigate the inherent modularity of
the object navigation task in Embodied AI. MOPA consists of four modules: (a)
an object detection module trained to identify objects from RGB images, (b) a
map building module to build a semantic map of the observed objects, (c) an
exploration module enabling the agent to explore the environment, and (d) a
navigation module to move to identified target objects. We show that we can
effectively reuse a pretrained PointGoal agent as the navigation model instead
of learning to navigate from scratch, thus saving time and compute. We also
compare various exploration strategies for MOPA and find that a simple uniform
strategy significantly outperforms more advanced exploration methods.
MOPA: Modular Object Navigation with PointGoal Agents. (arXiv:2304.03696v3 [cs.RO] UPDATED)
https://ift.tt/QlKUgBe
We propose a simple but effective modular approach MOPA (Modular ObjectNav with PointGoal agents) to systematically investigate the inherent modularity of the object navigation task in Embodied AI. MOPA consists of four modules: (a) an object detection module trained to identify objects from RGB images, (b) a map building module to build a semantic map of the observed objects, (c) an exploration module enabling the agent to explore the environment, and (d) a navigation module to move to identified target objects. We show that we can effectively reuse a pretrained PointGoal agent as the navigation model instead of learning to navigate from scratch, thus saving time and compute. We also compare various exploration strategies for MOPA and find that a simple uniform strategy significantly outperforms more advanced exploration methods.
via cs.RO updates on arXiv.org http://arxiv.org/