-
Multicollinearity-Aware Parameter-Free Strategy for Hyperspectral Band Selection: A Dependence Measures-Based Approach
Authors:
Dibyabha Deb,
Ujjwal Verma
Abstract:
Hyperspectral bands offer rich spectral and spatial information; however, their high dimensionality poses challenges for efficient processing. Band selection (BS) methods aim to extract a smaller subset of bands to reduce spectral redundancy. Existing approaches, such as ranking-based, clustering-based, and iterative methods, often suffer from issues like sensitivity to initialization, parameter t…
▽ More
Hyperspectral bands offer rich spectral and spatial information; however, their high dimensionality poses challenges for efficient processing. Band selection (BS) methods aim to extract a smaller subset of bands to reduce spectral redundancy. Existing approaches, such as ranking-based, clustering-based, and iterative methods, often suffer from issues like sensitivity to initialization, parameter tuning, and high computational cost. This work introduces a BS strategy integrating three dependence measures: Average Band Correlation (ABC) and Mutual Information (MI), and Variance Inflation Factor (VIF). ABC quantifies linear correlations between spectral bands, while MI measures uncertainty reduction relative to ground truth labels. To address multicollinearity and reduce the search space, the approach first applies a VIF-based pre-selection of spectral bands. Subsequently, a clustering algorithm is used to identify the optimal subset of bands based on the ABC and MI values. Unlike previous methods, this approach is completely parameter-free for hyperspectral band selection, eliminating the need for optimal parameter estimation. The proposed method is evaluated on four standard benchmark datasets: WHU-Hi-LongKou, Pavia University, Salinas, and Oil Spill datasets, and is compared to existing state-of-the-art approaches. There is significant overlap between the bands identified by our proposed method and those selected by other methods, indicating that our approach effectively captures the most relevant spectral features. Further, support vector machine (SVM) classification validates that VIF-driven pruning enhances classification by minimizing multicollinearity. Ablation studies confirm that combining ABC with MI yields robust, discriminative band subsets.
△ Less
Submitted 26 September, 2025;
originally announced September 2025.
-
Adversarial Robustness of Deep Learning Models for Inland Water Body Segmentation from SAR Images
Authors:
Siddharth Kothari,
Srinivasan Murali,
Sankalp Kothari,
Ujjwal Verma,
Jaya Sreevalsan-Nair
Abstract:
Inland water body segmentation from Synthetic Aperture Radar (SAR) images is an important task needed for several applications, such as flood mapping. While SAR sensors capture data in all-weather conditions as high-resolution images, differentiating water and water-like surfaces from SAR images is not straightforward. Inland water bodies, such as large river basins, have complex geometry, which a…
▽ More
Inland water body segmentation from Synthetic Aperture Radar (SAR) images is an important task needed for several applications, such as flood mapping. While SAR sensors capture data in all-weather conditions as high-resolution images, differentiating water and water-like surfaces from SAR images is not straightforward. Inland water bodies, such as large river basins, have complex geometry, which adds to the challenge of segmentation. U-Net is a widely used deep learning model for land-water segmentation of SAR images. In practice, manual annotation is often used to generate the corresponding water masks as ground truth. Manual annotation of the images is prone to label noise owing to data poisoning attacks, especially due to complex geometry. In this work, we simulate manual errors in the form of adversarial attacks on the U-Net model and study the robustness of the model to human errors in annotation. Our results indicate that U-Net can tolerate a certain level of corruption before its performance drops significantly. This finding highlights the crucial role that the quality of manual annotations plays in determining the effectiveness of the segmentation model. The code and the new dataset, along with adversarial examples for robust training, are publicly available. (GitHub link - https://github.com/GVCL/IWSeg-SAR-Poison.git)
△ Less
Submitted 6 May, 2025; v1 submitted 3 May, 2025;
originally announced May 2025.
-
Correlation-Based Band Selection for Hyperspectral Image Classification
Authors:
Dibyabha Deb,
Ujjwal Verma
Abstract:
Hyperspectral images offer extensive spectral information about ground objects across multiple spectral bands. However, the large volume of data can pose challenges during processing. Typically, adjacent bands in hyperspectral data are highly correlated, leading to the use of only a few selected bands for various applications. In this work, we present a correlation-based band selection approach fo…
▽ More
Hyperspectral images offer extensive spectral information about ground objects across multiple spectral bands. However, the large volume of data can pose challenges during processing. Typically, adjacent bands in hyperspectral data are highly correlated, leading to the use of only a few selected bands for various applications. In this work, we present a correlation-based band selection approach for hyperspectral image classification. Our approach calculates the average correlation between bands using correlation coefficients to identify the relationships among different bands. Afterward, we select a subset of bands by analyzing the average correlation and applying a threshold-based method. This allows us to isolate and retain bands that exhibit lower inter-band dependencies, ensuring that the selected bands provide diverse and non-redundant information. We evaluate our proposed approach on two standard benchmark datasets: Pavia University (PA) and Salinas Valley (SA), focusing on image classification tasks. The experimental results demonstrate that our method performs competitively with other standard band selection approaches.
△ Less
Submitted 24 January, 2025;
originally announced January 2025.
-
LapGSR: Laplacian Reconstructive Network for Guided Thermal Super-Resolution
Authors:
Aditya Kasliwal,
Ishaan Gakhar,
Aryan Kamani,
Pratinav Seth,
Ujjwal Verma
Abstract:
In the last few years, the fusion of multi-modal data has been widely studied for various applications such as robotics, gesture recognition, and autonomous navigation. Indeed, high-quality visual sensors are expensive, and consumer-grade sensors produce low-resolution images. Researchers have developed methods to combine RGB color images with non-visual data, such as thermal, to overcome this lim…
▽ More
In the last few years, the fusion of multi-modal data has been widely studied for various applications such as robotics, gesture recognition, and autonomous navigation. Indeed, high-quality visual sensors are expensive, and consumer-grade sensors produce low-resolution images. Researchers have developed methods to combine RGB color images with non-visual data, such as thermal, to overcome this limitation to improve resolution. Fusing multiple modalities to produce visually appealing, high-resolution images often requires dense models with millions of parameters and a heavy computational load, which is commonly attributed to the intricate architecture of the model.
We propose LapGSR, a multimodal, lightweight, generative model incorporating Laplacian image pyramids for guided thermal super-resolution. This approach uses a Laplacian Pyramid on RGB color images to extract vital edge information, which is then used to bypass heavy feature map computation in the higher layers of the model in tandem with a combined pixel and adversarial loss. LapGSR preserves the spatial and structural details of the image while also being efficient and compact. This results in a model with significantly fewer parameters than other SOTA models while demonstrating excellent results on two cross-domain datasets viz. ULB17-VT and VGTSR datasets.
△ Less
Submitted 12 November, 2024;
originally announced November 2024.
