Summary of Kolmogorov-arnold Network For Satellite Image Classification in Remote Sensing, by Minjong Cheon
Kolmogorov-Arnold Network for Satellite Image Classification in Remote Sensing
by Minjong Cheon
First submitted to arxiv on: 2 Jun 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Data Analysis, Statistics and Probability (physics.data-an)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This research proposes an innovative approach to integrating the Kolmogorov-Arnold Network (KAN) with various pre-trained Convolutional Neural Network (CNN) models for remote sensing scene classification tasks using the EuroSAT dataset. The novel methodology, KCN, aims to replace traditional Multi-Layer Perceptrons (MLPs) with KAN to enhance classification performance. Multiple CNN-based models, including VGG16, MobileNetV2, EfficientNet, ConvNeXt, ResNet101, and Vision Transformer (ViT), are employed and evaluated in conjunction with KAN. The results demonstrate that KAN achieves high accuracy with fewer training epochs and parameters, with ConvNeXt paired with KAN showing the best performance. This research provides a robust testbed for investigating whether KAN is suitable for remote sensing classification tasks, offering a promising alternative for efficient image analysis. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about using a new type of artificial intelligence called the Kolmogorov-Arnold Network (KAN) to help computers recognize pictures taken from space. The researchers combined KAN with different types of computer vision models and tested them on a large dataset of satellite images. They found that KAN worked really well, especially when used with one specific type of model called ConvNeXt. This is important because it could help us analyze these images more quickly and accurately, which has many potential applications in fields like environmental monitoring and disaster response. |
Keywords
» Artificial intelligence » Classification » Cnn » Neural network » Vision transformer » Vit
