Data Efficient Human Intention Prediction: Leveraging Neural Network Verification and Expert Guidance

TL;DR

This paper introduces an iterative adversarial data augmentation framework that uses neural network verification and expert guidance to improve human intention prediction with limited data, enhancing robustness and accuracy.

Contribution

It presents a novel IADA framework combining neural network verification and expert input to augment training data for better intention prediction in HRC.

Findings

Improved prediction robustness over existing methods

Effective augmentation of scarce data with challenging samples

Enhanced model accuracy in human intention prediction

Abstract

Predicting human intention is critical to facilitating safe and efficient human-robot collaboration (HRC). However, it is challenging to build data-driven models for human intention prediction. One major challenge is due to the diversity and noise in human motion data. It is expensive to collect a massive motion dataset that comprehensively covers all possible scenarios, which leads to the scarcity of human motion data in certain scenarios, and therefore, causes difficulties in constructing robust and reliable intention predictors. To address the challenge, this paper proposes an iterative adversarial data augmentation (IADA) framework to learn neural network models from an insufficient amount of training data. The method uses neural network verification to identify the most "confusing" input samples and leverages expert guidance to safely and iteratively augment the training data with these samples. The proposed framework is applied to collected human datasets. The experiments demonstrate that our method can achieve more robust and accurate prediction performance compared to existing training methods.