Private Multiparty Perception for Navigation

TL;DR

This paper presents a privacy-preserving multiparty perception framework that enables navigation through complex environments using multiple cameras while preventing privacy breaches, scalable to many viewpoints.

Contribution

It introduces a novel multiview scene representation method that ensures privacy during navigation, with proven privacy guarantees and scalability to multiple camera inputs.

Findings

Enables navigation in cluttered environments with privacy preservation.

Scales effectively to an arbitrary number of camera viewpoints.

Demonstrates successful navigation while maintaining privacy on a new dataset.

Abstract

We introduce a framework for navigating through cluttered environments by connecting multiple cameras together while simultaneously preserving privacy. Occlusions and obstacles in large environments are often challenging situations for navigation agents because the environment is not fully observable from a single camera view. Given multiple camera views of an environment, our approach learns to produce a multiview scene representation that can only be used for navigation, provably preventing one party from inferring anything beyond the output task. On a new navigation dataset that we will publicly release, experiments show that private multiparty representations allow navigation through complex scenes and around obstacles while jointly preserving privacy. Our approach scales to an arbitrary number of camera viewpoints. We believe developing visual representations that preserve privacy is increasingly important for many applications such as navigation.