Multiple Thinking Achieving Meta-Ability Decoupling for Object Navigation

TL;DR

This paper introduces a meta-ability decoupling paradigm and a multiple thinking model to enhance object navigation by enabling different meta-abilities to collaborate and evolve, resulting in improved performance and interpretability.

Contribution

The paper presents a novel MAD paradigm and a multiple thinking model that decouple and integrate meta-abilities for object navigation, advancing interpretability and performance.

Findings

Outperforms SOTA on AI2-Thor and RoboTHOR datasets

Effective decoupling of meta-abilities from input, encoding, and reward

Introduces a new interpretability system for object navigation

Abstract

We propose a meta-ability decoupling (MAD) paradigm, which brings together various object navigation methods in an architecture system, allowing them to mutually enhance each other and evolve together. Based on the MAD paradigm, we design a multiple thinking (MT) model that leverages distinct thinking to abstract various meta-abilities. Our method decouples meta-abilities from three aspects: input, encoding, and reward while employing the multiple thinking collaboration (MTC) module to promote mutual cooperation between thinking. MAD introduces a novel qualitative and quantitative interpretability system for object navigation. Through extensive experiments on AI2-Thor and RoboTHOR, we demonstrate that our method outperforms state-of-the-art (SOTA) methods on both typical and zero-shot object navigation tasks.