Skip to main content

Research Repository

Advanced Search

Improving Generalization for Hyperspectral Image Classification: The Impact of Disjoint Sampling on Deep Models

Ahmad, Muhammad; Mazzara, Manuel; Distefano, Salvatore; Khan, Adil Mehmood; Altuwaijri, Hamad Ahmed

Authors

Muhammad Ahmad

Manuel Mazzara

Salvatore Distefano

Hamad Ahmed Altuwaijri



Abstract

Disjoint sampling is critical for rigorous and unbiased evaluation of state-of-the-art (SOTA) models e.g., Attention Graph and Vision Transformer. When training, validation, and test sets overlap or share data, it introduces a bias that inflates performance metrics and prevents accurate assessment of a model’s true ability to generalize to new examples. This paper presents an innovative disjoint sampling approach for training SOTA models for the Hyperspectral Image Classification (HSIC). By separating training, validation, and test data without overlap, the proposed method facilitates a fairer evaluation of how well a model can classify pixels it was not exposed to during training or validation. Experiments demonstrate the approach significantly improves a model’s generalization compared to alternatives that include training and validation data in test data (A trivial approach involves testing the model on the entire Hyperspectral dataset to generate the ground truth maps. This approach produces higher accuracy but ultimately results in low generalization performance). Disjoint sampling eliminates data leakage between sets and provides reliable metrics for benchmarking progress in HSIC. Disjoint sampling is critical for advancing SOTA models and their real-world application to large-scale land mapping with Hyperspectral sensors. Overall, with the disjoint test set, the performance of the deep models achieves 96.36% accuracy on Indian Pines data, 99.73% on Pavia University data, 98.29% on University of Houston data, 99.43% on Botswana data, and 99.88% on Salinas data.

Citation

Ahmad, M., Mazzara, M., Distefano, S., Khan, A. M., & Altuwaijri, H. A. (2024). Improving Generalization for Hyperspectral Image Classification: The Impact of Disjoint Sampling on Deep Models. Computers, Materials & Continua, 81(1), 503-532. https://doi.org/10.32604/cmc.2024.056318

Journal Article Type Article
Acceptance Date Aug 11, 2024
Online Publication Date Oct 15, 2024
Publication Date Oct 15, 2024
Deposit Date Oct 21, 2024
Publicly Available Date Oct 21, 2024
Journal Computers, Materials & Continua
Print ISSN 1546-2218
Electronic ISSN 1546-2226
Publisher Tech Science Press
Peer Reviewed Peer Reviewed
Volume 81
Issue 1
Pages 503-532
DOI https://doi.org/10.32604/cmc.2024.056318
Keywords Hyperspectral image classification; Disjoint sampling; Graph CNN; Spatial-spectral transformer
Public URL https://hull-repository.worktribe.com/output/4868888

Files

Published article (4.1 Mb)
PDF

Publisher Licence URL
http://creativecommons.org/licenses/by/4.0

Copyright Statement
Copyright © 2024 The Authors. Published by Tech Science Press.
This work is licensed under a Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited




You might also like



Downloadable Citations