Self-supervised Transparent Liquid Segmentation for Robotic Pouring

TL;DR

This paper introduces a novel self-supervised segmentation method for transparent liquids in RGB images, enabling robotic pouring tasks without manual annotations by leveraging generative models and background subtraction.

Contribution

It presents a new segmentation pipeline that uses generative models trained on unpaired data to accurately segment transparent liquids without manual labels.

Findings

Accurately segments transparent liquids in static RGB images.

Enables robotic pouring by perceiving liquid height.

Operates without manual annotations or liquid heating.

Abstract

Liquid state estimation is important for robotics tasks such as pouring; however, estimating the state of transparent liquids is a challenging problem. We propose a novel segmentation pipeline that can segment transparent liquids such as water from a static, RGB image without requiring any manual annotations or heating of the liquid for training. Instead, we use a generative model that is capable of translating images of colored liquids into synthetically generated transparent liquid images, trained only on an unpaired dataset of colored and transparent liquid images. Segmentation labels of colored liquids are obtained automatically using background subtraction. Our experiments show that we are able to accurately predict a segmentation mask for transparent liquids without requiring any manual annotations. We demonstrate the utility of transparent liquid segmentation in a robotic pouring task that controls pouring by perceiving the liquid height in a transparent cup. Accompanying video and supplementary materials can be found