WaveMamba: Enhancing Hyperspectral Image Classification by Integrating Wavelet Transformation and Spatial-Spectral Mamba Architecture
WaveMamba, a novel approach integrating wavelet transformation with the spatial-spectral Mamba architecture, enhances hyperspectral image classification accuracy by effectively capturing both local texture patterns and global contextual relationships.
Hyperspectral Imaging and Few-Shot Learning for Grain Quality Assessment
Few-shot learning (FSL) combined with hyperspectral imaging (HSI) offers a practical and accurate solution for grain quality assessment, particularly in scenarios with limited labeled data, achieving comparable results to fully trained classifiers while requiring significantly fewer training samples.
Improving Hyperspectral Image Classification Generalization with a Spectral-spatial Axial Aggregation Transformer
Deep learning models often exhibit overly optimistic performance in hyperspectral image classification due to information sharing between training and test datasets, and this paper proposes a novel transformer-based model, SaaFormer, to enhance generalization ability by emphasizing spectral feature extraction and mitigating data leakage issues through alternative sampling methods.
Enhancing Hyperspectral Image Prediction Using Contrastive Learning with Limited Labeled Data
Contrastive learning, a self-supervised learning technique, can significantly improve the performance of hyperspectral image classification, especially in scenarios with limited labeled training data.
웨이블릿 기반 콜모고로프-아놀드 네트워크를 이용한 초분광 영상 분류
본 논문에서는 웨이블릿 함수를 학습 가능한 활성화 함수로 활용하는 Wav-KAN(Wavelet-based Kolmogorov-Arnold Network)이 초분광 영상 분류 작업에서 MLP 및 Spline-KAN보다 우수한 성능을 보임을 입증합니다.
Wavelet-based Kolmogorov-Arnold Network for Hyperspectral Image Classification: A Comparative Study on Benchmark Datasets
This research paper presents Wav-KAN, a novel wavelet-based Kolmogorov-Arnold Network architecture, for enhanced hyperspectral image classification, demonstrating its superior performance over traditional MLPs and Spline-KAN on benchmark datasets.
Interval Group Spatial-Spectral Mamba: A Lightweight Framework for Enhanced Hyperspectral Image Classification using Selective State Space Models
This paper introduces IGroupSS-Mamba, a novel deep learning framework for hyperspectral image classification that leverages the strengths of Selective State Space Models (SSMs) in a computationally efficient manner to achieve state-of-the-art classification accuracy.
Selective Transformer (SFormer): A Novel Deep Learning Model for Enhanced Hyperspectral Image Classification
The SFormer model leverages selective attention mechanisms to dynamically adapt receptive fields and prioritize relevant spatial-spectral information, leading to improved accuracy in hyperspectral image classification compared to traditional CNNs and existing transformer-based methods.
Enhancing Hyperspectral Image Classification with a Novel CNN-Transformer Approach Leveraging Gate-Shift-Fuse Mechanisms
A novel CNN-Transformer approach with Gate-Shift-Fuse (GSF) mechanisms is proposed to effectively extract local and global spatial-spectral features for enhanced hyperspectral image classification.
Attentional Fusion of 3D Swin Transformer and Spatial-Spectral Transformer for Robust Hyperspectral Image Classification
The proposed method introduces an attentional fusion of 3D Swin Transformer and Spatial-Spectral Transformer to significantly enhance the classification performance of Hyperspectral Images by leveraging the complementary strengths of hierarchical attention, window-based processing, and long-range dependency modeling.
© 2024 by Linnk AI