Active Learning of Robot Vision Using Adaptive Path Planning

TL;DR

This paper introduces an adaptive path planning framework for robotic vision that reduces human labeling effort by combining human and pseudo labels, achieving near-supervised performance in semantic terrain monitoring.

Contribution

It presents a novel adaptive planning approach that efficiently collects training data, significantly lowering human labeling needs while maintaining high segmentation accuracy.

Findings

Achieves segmentation performance close to fully supervised methods.

Reduces human labeling effort substantially.

Outperforms purely self-supervised approaches.

Abstract

Robots need robust and flexible vision systems to perceive and reason about their environments beyond geometry. Most of such systems build upon deep learning approaches. As autonomous robots are commonly deployed in initially unknown environments, pre-training on static datasets cannot always capture the variety of domains and limits the robot's vision performance during missions. Recently, self-supervised as well as fully supervised active learning methods emerged to improve robotic vision. These approaches rely on large in-domain pre-training datasets or require substantial human labelling effort. To address these issues, we present a recent adaptive planning framework for efficient training data collection to substantially reduce human labelling requirements in semantic terrain monitoring missions. To this end, we combine high-quality human labels with automatically generated pseudo labels. Experimental results show that the framework reaches segmentation performance close to fully supervised approaches with drastically reduced human labelling effort while outperforming purely self-supervised approaches. We discuss the advantages and limitations of current methods and outline valuable future research avenues towards more robust and flexible robotic vision systems in unknown environments.