Supervised Bayesian Specification Inference from Demonstrations

TL;DR

This paper introduces a Bayesian approach to infer task specifications expressed as temporal logic formulas from demonstrations, enabling early assessment of task correctness with high accuracy.

Contribution

It proposes a probabilistic model that combines priors and likelihoods for inferring temporal logic specifications from demonstrations, advancing learning from demonstrations methods.

Findings

Over 90% similarity between inferred and true specifications

Effective in both synthetic and real-world tasks

Demonstrates high accuracy in specification inference

Abstract

When observing task demonstrations, human apprentices are able to identify whether a given task is executed correctly long before they gain expertise in actually performing that task. Prior research into learning from demonstrations (LfD) has failed to capture this notion of the acceptability of a task's execution; meanwhile, temporal logics provide a flexible language for expressing task specifications. Inspired by this, we present Bayesian specification inference, a probabilistic model for inferring task specification as a temporal logic formula. We incorporate methods from probabilistic programming to define our priors, along with a domain-independent likelihood function to enable sampling-based inference. We demonstrate the efficacy of our model for inferring specifications, with over 90% similarity observed between the inferred specification and the ground truth, both within a synthetic domain and during a real-world table setting task.