Adaptive spectral-spatial feature fusion network for hyperspectral image classification using limited training samples
Abstract: Highlights•A model that can perform excellent classification performance under the condition of limited training sample size is proposed.•Both near-band inter-relationships and far-band inter-relationships are considered in the spectral feature extraction process.•Multiscale spatial feature extraction via multiscale-share inception block.•Adaptive fusion of spectral and spatial features.
Loading