Kolmogorov-Arnold Network for Satellite Image Classification in Remote Sensing

Minjong Cheon

2024-06-02

Abstract:In this research, we propose the first approach for integrating the Kolmogorov-Arnold Network (KAN) with various pre-trained Convolutional Neural Network (CNN) models for remote sensing (RS) scene classification tasks using the EuroSAT dataset. Our novel methodology, named KCN, aims to replace traditional Multi-Layer Perceptrons (MLPs) with KAN to enhance classification performance. We employed multiple CNN-based models, including VGG16, MobileNetV2, EfficientNet, ConvNeXt, ResNet101, and Vision Transformer (ViT), and evaluated their performance when paired with KAN. Our experiments demonstrated that KAN achieved high accuracy with fewer training epochs and parameters. Specifically, ConvNeXt paired with KAN showed the best performance, achieving 94% accuracy in the first epoch, which increased to 96% and remained consistent across subsequent epochs. The results indicated that KAN and MLP both achieved similar accuracy, with KAN performing slightly better in later epochs. By utilizing the EuroSAT dataset, we provided a robust testbed to investigate whether KAN is suitable for remote sensing classification tasks. Given that KAN is a novel algorithm, there is substantial capacity for further development and optimization, suggesting that KCN offers a promising alternative for efficient image analysis in the RS field.

Computer Vision and Pattern Recognition,Artificial Intelligence,Data Analysis, Statistics and Probability

What problem does this paper attempt to address?

The problem this paper attempts to address is: In the task of Remote Sensing (RS) scene classification, how to use the Kolmogorov-Arnold Network (KAN) to replace traditional Multi-Layer Perceptrons (MLPs) to improve classification performance and reduce the number of parameters and training time required.

Specifically, the paper proposes a new method—Kolmogorov-Arnold Network with Convolutional Neural Networks (KCN), which combines KAN with various pre-trained convolutional neural networks (such as VGG16, MobileNetV2, EfficientNet, ConvNeXt, ResNet101, and Vision Transformer) for satellite image classification tasks. Through experimental validation, KAN can achieve high accuracy with fewer training epochs and parameters, performing best when combined with ConvNeXt, reaching 94% accuracy in the first training epoch and maintaining stability in subsequent epochs.

The main contributions of the paper include:
1. Utilizing the Kolmogorov-Arnold Network (KAN) architecture to replace traditional MLPs for remote sensing scene classification tasks.
2. Determining the most suitable model to pair with the KAN framework by comparing various pre-trained CNN and Vision Transformer models.
3. Evaluating the efficiency of KAN through comparative experiments with different parameter configurations, demonstrating its potential application value in multiple fields.