HybridGBN-SR: A Deep 3D/2D Genome Graph-Based Network for Hyperspectral Image Classification

Abstract

The successful application of deep learning approaches in remote sensing image classification requires large hyperspectral image (HSI) datasets to learn discriminative spectral–spatial features simultaneously. To date, the HSI datasets available for image classification are relatively small to train deep learning methods. This study proposes a deep 3D/2D genome graph-based network (abbreviated as HybridGBN-SR) that is computationally efficient and not prone to overfitting even with extremely few training sample data. At the feature extraction level, the HybridGBN-SR utilizes the three-dimensional (3D) and two-dimensional (2D) Genoblocks trained using very few samples while improving HSI classification accuracy. The design of a Genoblock is based on a biological genome graph. From the experimental results, the study shows that our model achieves better classification accuracy than the compared state-of-the-art methods over the three publicly available HSI benchmarking datasets such as the Indian Pines (IP), the University of Pavia (UP), and the Salinas Scene (SA). For instance, using only 5% labeled data for training in IP, and 1% in UP and SA, the overall classification accuracy of the proposed HybridGBN-SR is 97.42%, 97.85%, and 99.34%, respectively, which is better than the compared state-of-the-art methods.

Advancing Hyperspectral Image Analysis with CTNet: An Approach with the Fusion of Spatial and Spectral Features

“…where I_N = Normalized pixel value rounded to the nearest integer and I_N(i, j, c ) = normalized intensity value of the pixel at position (i, j) in channel c. (6) Construct the final RGB image using the normalized and rounded values in each channel as follows.…”

“…Further, several research studies have utilized joint spectral and spatial features to improve the classification [ 5 ]. Vision transformers (ViTs) have recently been proposed to provide long-range dependency on spatial and spectral features for the classification of land objects [ 6 ].…”

Improving Feature Learning in Remote Sensing Images Using an Integrated Deep Multi-Scale 3D/2D Convolutional Network

“…The usage of non-identity multi-residual connections drastically reduces the challenge of gradient disappearance in the MiCB network [15] while replacing the traditional 2D-CNN with the 2D depthwise separable convolutional layers promotes the reduction in network parameters and prevents overfitting as the model structure deepens. Lastly, the utilization of multi-scale kernels enhances the extraction of abundant contextual features [23][24][25][26]. At the bottom of the MiCB network architecture, we utilized the 3D convolution to extract HSI spectral-spatial features, as shown in Figure 8.…”

“…Figure 13 depicts the framework for multi-scale feature learning with kernels of various sizes to identify a broader range of significant characteristics [23,[25][26][27]. The red, aqua, and blue boxes are distinct convolutional filters used to discover hidden characteristics.…”

“…Tinega et al [1], developed a GGBN model that used the biological genome concept to combine 3D-CNN and 2D-CNN and residual connections in a network structure [1]. Among other researchers who utilized a mixture of 3D-CNN and 2D-CNN in their network structure, and residual connections to produce the state-of-the-art models include zhao et al [22], and Tinega et al [23].…”