Self-Distillation-Based Polarimetric Image Classification with Noisy and Sparse Labels
Open Access
- 15 December 2023
- journal article
- research article
- Published by MDPI AG in Remote Sensing
- Vol. 15 (24) , 5751
Abstract
Polarimetric synthetic aperture radar (PolSAR) image classification, a field crucial in remote sensing, faces significant challenges due to the intricate expertise required for accurate annotation, leading to susceptibility to labeling inaccuracies. Compounding this challenge are the constraints posed by limited labeled samples and the perennial issue of class imbalance inherent in PolSAR image classification. Our research objectives are to address these challenges by developing a novel label correction mechanism, implementing self-distillation-based contrastive learning, and introducing a sample rebalancing loss function. To address the quandary of noisy labels, we proffer a novel label correction mechanism that capitalizes on inherent sample similarities to rectify erroneously labeled instances. In parallel, to mitigate the limitation of sparsely labeled data, this study delves into self-distillation-based contrastive learning, harnessing sample affinities for nuanced feature extraction. Moreover, we introduce a sample rebalancing loss function that adjusts class weights and augments data for small classes. Through extensive experiments on four benchmark PolSAR images, our approach demonstrates its effectiveness in addressing label inaccuracies, limited samples, and class imbalance. Through extensive experiments on four benchmark PolSAR images, our research substantiates the robustness of our proposed methodology, particularly in rectifying label discrepancies in contexts marked by sample paucity and imbalance. The empirical findings illuminate the superior efficacy of our approach, positioning it at the forefront of state-of-the-art PolSAR classification techniques.Keywords
Funding Information
- National Key R&D Program of China (2022YFA1003800, 42201394, PCL2023AS1-2, QCYRCXM-2022-30)
- National Natural Science Foundation of China (2022YFA1003800, 42201394, PCL2023AS1-2, QCYRCXM-2022-30)
- Major Key Project of Peng Cheng Laboratory (2022YFA1003800, 42201394, PCL2023AS1-2, QCYRCXM-2022-30)
- Qinchuangyuan High-level Innovation and Entrepreneurial Talent Program (2022YFA1003800, 42201394, PCL2023AS1-2, QCYRCXM-2022-30)
This publication has 40 references indexed in Scilit:
- Robust Semisupervised Classification for PolSAR Image With Noisy LabelsIEEE Transactions on Geoscience and Remote Sensing, 2017
- Polarimetric SAR Image Classification Using Deep Convolutional Neural NetworksIEEE Geoscience and Remote Sensing Letters, 2016
- PolSAR image classification using a semi-supervised classifier based on hypergraph learningRemote Sensing Letters, 2014
- Unsupervised Land Cover/Land Use Classification Using PolSAR Imagery Based on Scattering SimilarityIEEE Transactions on Geoscience and Remote Sensing, 2012
- Laplacian Eigenmaps-Based Polarimetric Dimensionality Reduction for SAR Image ClassificationIEEE Transactions on Geoscience and Remote Sensing, 2011
- Unsupervised Polarimetric SAR Image Segmentation and Classification Using Region Growing With Edge PenaltyIEEE Transactions on Geoscience and Remote Sensing, 2011
- Region-Based Classification of Polarimetric SAR Images Using Wishart MRFIEEE Geoscience and Remote Sensing Letters, 2008
- A Tutorial on the Cross-Entropy MethodAnnals of Operations Research, 2005
- A review of target decomposition theorems in radar polarimetryIEEE Transactions on Geoscience and Remote Sensing, 1996
- Unsupervised classification of scattering behavior using radar polarimetry dataIEEE Transactions on Geoscience and Remote Sensing, 1989