Safety Certification in the Latent space using Control Barrier Functions and World Models

TL;DR

This paper presents a semi-supervised approach for safe visuomotor control by learning control barrier certificates in the latent space of world models, reducing the need for extensive safety data.

Contribution

It introduces a novel framework that combines control barrier certificates with world models and vision transformers for scalable, data-efficient safety in visual control tasks.

Findings

Successfully synthesizes safe policies with limited labeled data

Leverages vision transformers for accurate latent dynamics prediction

Demonstrates improved safety performance in visual control environments

Abstract

Synthesising safe controllers from visual data typically requires extensive supervised labelling of safety-critical data, which is often impractical in real-world settings. Recent advances in world models enable reliable prediction in latent spaces, opening new avenues for scalable and data-efficient safe control. In this work, we introduce a semi-supervised framework that leverages control barrier certificates (CBCs) learned in the latent space of a world model to synthesise safe visuomotor policies. Our approach jointly learns a neural barrier function and a safe controller using limited labelled data, while exploiting the predictive power of modern vision transformers for latent dynamics modelling.