Abstract:Recently, researchers have shown the powerful ability of deep methods with multilayers to extract high-level features and to obtain better performance for hyperspectral image classification. However, a common problem of traditional deep models is that the learned deep models might be suboptimal because of the limited number of training samples, especially for the image with large intraclass variance and low interclass variance. In this paper, novel convolutional neural networks (CNNs) with multiscale convolution (MS-CNNs) are proposed to address this problem by extracting deep multiscale features from the hyperspectral image. Moreover, deep metrics usually accompany with MS-CNNs to improve the representational ability for the hyperspectral image. However, the usual metric learning would make the metric parameters in the learned model tend to behave similarly. This similarity leads to obvious model’s redundancy and, thus, shows negative effects on the description ability of the deep metrics. Traditionally, determinantal point process (DPP) priors, which encourage the learned factors to repulse from one another, can be imposed over these factors to diversify them. Taking advantage of both the MS-CNNs and DPP-based diversity-promoting deep metrics, this paper develops a CNN with multiscale convolution and diversified metric to obtain discriminative features for hyperspectral image classification. Experiments are conducted over four real-world hyperspectral image data sets to show the effectiveness and applicability of the proposed method. Experimental results show that our method is better than original deep models and can produce comparable or even better classification performance in different hyperspectral image data sets with respect to spectral and spectral–spatial features.

Cyclic CNN: Image Classification With Multiscale and Multilocation Contexts

Transform-Invariant Convolutional Neural Networks for Image Classification and Search

Rotation-Invariant Feature Learning Via Convolutional Neural Network with Cyclic Polar Coordinates Convolutional Layer

Hyperspectral Classification Via Spatial Context Exploration with Multi-Scale CNN

Consecutive multiscale feature learning-based image classification model

Coarse-to-fine Trained Multi-Scale Convolutional Neural Networks for Image Classification.

CDTNet: Improved Image Classification Method Using Standard, Dilated and Transposed Convolutions

Scene Classification Based on Multiscale Convolutional Neural Network.

A cnn-gcn framework for multi-label aerial image scene classification

CCTSS: the Combination of CNN and Transformer with Shared Sublayer for Detection and Classification

Multi-Scale Dense Networks for Resource Efficient Image Classification

Multi-Scale Dense Convolutional Networks for Efficient Prediction.

Multiscale Low-Frequency Memory Network for Improved Feature Extraction in Convolutional Neural Networks

A CNN With Multiscale Convolution and Diversified Metric for Hyperspectral Image Classification

A Compact Deep Neural Network for Single Image Super-Resolution

LiteCCLKNet: A Lightweight Criss‐cross Large Kernel Convolutional Neural Network for Hyperspectral Image Classification

Multi-task Cascade Deep Convolutional Neural Networks for Large-Scale Commodity Recognition

Learning Contextual Dependencies with Convolutional Hierarchical Recurrent Neural Networks

Circulant Binary Convolutional Networks for Object Recognition.

CS-CNN: Enabling Robust and Efficient Convolutional Neural Networks Inference for Internet-of-Things Applications

Tensor Neural Networks Via Circulant Convolution