Supervised Bayesian Specification Inference from Demonstrations
This addresses the challenge of learning task acceptability from demonstrations for robotics and AI systems, representing a novel method for a known bottleneck.
The paper tackled the problem of inferring task specifications from demonstrations by introducing Bayesian specification inference, a probabilistic model that uses temporal logic formulas, achieving over 90% similarity to ground truth in synthetic and real-world tasks.
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.