A query-driven twin network framework with optimization-based meta-learning for few-shot hyperspectral image classification

Jian Zhu

doi:10.1016/j.patcog.2025.112331

A query-driven twin network framework with optimization-based meta-learning for few-shot hyperspectral image classification

2025-08-31

0

PRE

AI

DOI：10.1016/j.patcog.2025.112331

Original

Original request for help

Save

Abstract

En 中文

• Designed a lightweight spectral spatial-attention residual network (SSARN) for efficient feature extraction. • Proposed query-loss-only meta-learning (QLOML) algorithm which using twin networks to separate meta learning. • Two meta-task generation strategies enable extensive evaluation on three datasets, demonstrating high effectiveness.

A query-driven twin network framework with optimization-based meta-learning for few-shot hyperspectral image classification

Abstract

Journal

Organization