IPPON: Common Sense Guided Informative Path Planning for Object Goal Navigation

1Robotic Systems Lab, ETH Zurich 2Google DeepMind

The robot finds eight different objects in succession: a toy elephant, a screwdriver, a microwave oven, a coffee machine, a sofa, a plant, a chair, and a TV. We send the next target immediately when the robot locates the current object of interest and the robot can reuse the probability map and ESDF for an efficient search.

Abstract

Navigating efficiently to an object in an unexplored environment is a critical skill for general-purpose intelligent robots. Recent approaches to this object goal navigation problem have embraced a modular strategy, integrating classical exploration algorithms—notably frontier exploration—with a learned semantic mapping/exploration module. This paper introduces a novel informative path planning and 3D object probability mapping approach. The mapping module computes the probability of the object of interest through semantic segmentation and a Bayes filter. Additionally, it stores probabilities for common objects, which semantically guides the exploration based on common sense priors from a large language model. The planner terminates when the current viewpoint captures enough voxels identified with high confidence as the object of interest. Although our planner follows a zero-shot approach, it achieves state-of-the-art performance as measured by the Success weighted by Path Length (SPL) and Soft SPL in the Habitat ObjectNav Challenge 2023, outperforming other works by more than 20%. Furthermore, we validate its effectiveness on real robots.

Video

Approach

Ippon pipleline figure.

The pipeline of IPPON consists of three main components: 3D object probability mapping using the Bayes Filter, common sense reasoning that provides a proximity map for the OOI, and informative path planning based on the probability map combined with the proximity map. IPPON uses voxblox to compute the Euclidean signed distance field (ESDF) online for traversability estimation.

Habitat ObjectNav Evaluation

Evaluation results on the Habitat ObjectNav 2023 Challenge (standard phase)
Method SPL ↑ Soft SPL ↑ Success ↑
Host team 0.05 0.27 0.12
Auxiliary RL 0.10 0.31 0.18
ICanFly 0.26 0.37 0.43
SkillFusion 0.28 0.34 0.53
SkillTron 0.28 0.36 0.59
IPPON (ours) 0.34 0.46 0.54

Object Proximity from LLM Reasoning

Prompt

Response