Life is not black and white -- Combining Semi-Supervised Learning with fuzzy labels

TL;DR

This paper explores integrating fuzzy labels into semi-supervised learning to reduce annotation costs and improve consistency, addressing variability in annotations often overlooked in current methods.

Contribution

It introduces a novel approach to incorporate fuzzy labels into semi-supervised learning, demonstrating potential for lower costs and higher annotation consistency.

Findings

Fuzzy labels can be effectively integrated into semi-supervised learning.

Incorporating fuzzy labels may reduce annotation costs.

The approach improves label consistency across datasets.

Abstract

The required amount of labeled data is one of the biggest issues in deep learning. Semi-Supervised Learning can potentially solve this issue by using additional unlabeled data. However, many datasets suffer from variability in the annotations. The aggregated labels from these annotation are not consistent between different annotators and thus are considered fuzzy. These fuzzy labels are often not considered by Semi-Supervised Learning. This leads either to an inferior performance or to higher initial annotation costs in the complete machine learning development cycle. We envision the incorporation of fuzzy labels into Semi-Supervised Learning and give a proof-of-concept of the potential lower costs and higher consistency in the complete development cycle. As part of our concept, we discuss current limitations, futures research opportunities and potential broad impacts.