FADC-Net: Frequency-Aware Adaptive Dilation Convolution Network With Spatial-Spectral Enhancement for Hyperspectral Image Classification

Accurate hyperspectral image (HSI) classification requires jointly modeling fine-grained spatial textures, dense spectral signatures, and frequency cues, yet most existing approaches still rely on static frequency processing and fixed receptive fields, which limits their adaptability across scenes....

Ausführliche Beschreibung

Bibliografische Detailangaben

1. Verfasser:	Yufan Chen
Format:	Artigo
Sprache:	Inglês
Veröffentlicht:	IEEE 2026-01-01
Schriftenreihe:	IEEE Access
Schlagworte:	Hyperspectral image classification deep learning frequency-aware learning adaptive dilation convolution spatial-spectral feature extraction Fourier transform
Online-Zugang:	https://ieeexplore.ieee.org/document/11471755/
Tags:	Tag hinzufügen Keine Tags, Fügen Sie das erste Tag hinzu!