An assessment of data-centric methods for label noise identification in remote sensing data sets

2026-03-17

AI Summary

Label noise is prevalent in remote sensing datasets and significantly impairs the generalization capability of deep learning models; however, existing literature lacks a systematic evaluation of data-centric approaches for handling such noise in this domain. This work presents the first comprehensive comparison of three data-centric label noise identification methods within remote sensing, conducting quantitative experiments on two benchmark datasets with synthetically injected noise of varying types and intensities (10%–70%). The results reveal distinct performance patterns across methods in terms of noise identification accuracy and downstream task improvement, clarifying their respective applicability scenarios. These findings provide empirical guidance for selecting denoising strategies in remote sensing applications and highlight critical gaps for future research.

Technology Category

• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Computer Vision: Remote Sensing / Geospatial AI
• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)

Application Category

• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Label noise in the sense of incorrect labels is present in many real-world data sets and is known to severely limit the generalizability of deep learning models. In the field of remote sensing, however, automated treatment of label noise in data sets has received little attention to date. In particular, there is a lack of systematic analysis of the performance of data-centric methods that not only cope with label noise but also explicitly identify and isolate noisy labels. In this paper, we examine three such methods and evaluate their behavior under different label noise assumptions. To do this, we inject different types of label noise with noise levels ranging from 10 to 70% into two benchmark data sets, followed by an analysis of how well the selected methods filter the label noise and how this affects task performances. With our analyses, we clearly prove the value of data-centric methods for both parts - label noise identification and task performance improvements. Our analyses provide insights into which method is the best choice depending on the setting and objective. Finally, we show in which areas there is still a need for research in the transfer of data-centric label noise methods to remote sensing data. As such, our work is a step forward in bridging the methodological establishment of data-centric label noise methods and their usage in practical settings in the remote sensing domain.

Problem

• label noise
• remote sensing
• data-centric methods
• noise identification
• deep learning

Innovation

• data-centric methods
• label noise identification
• remote sensing
• noise robustness
• benchmark evaluation