Abstract:Hyperspectral anomaly detection, which aims to search and localize potential targets, is a research area with extensive application prospects and profound implications. In recent years, the emergence of unsupervised and self-supervised deep learning for image reconstruction has provided inspiring solutions for hyperspectral anomaly detection. However, due to sensor-induced and environmental effects, the full-image detection networks inevitably reconstruct anomalies along with the background. Existing detectors indirectly mitigate anomaly reconstruction by imposing constraints on hidden features or loss functions, but they provide unsatisfactory performance in large target detection scenarios. This work straightforwardly addresses this issue from the input source, i.e., introducing the concept of masked autoencoders (MAEs) into fully convolutional networks and further developing a sparse-inspired masked convolutional network (SMCNet) consisting of three mutually supportive components: 1) a hierarchical encoder; 2) a sparse projection layer; and 3) a hierarchical decoder. The encoder employs an adaptive potential anomaly masking strategy, leveraging sparse convolution for extracting multidimensional features of the remaining background. Meanwhile, a sparse-guided projection layer is created by discarding the positional embedding technique to populate the uncoded region and guide the background recovery without introducing anomalies. Finally, the decoder couples the hierarchical structure and a hybrid attention mechanism (local-middle–global and spatial-spectral) to refine the background during image recovery, whereas anomalies in the residual map are highlighted. Extensive experiments using ten typical competitors on six different types of datasets validate the effectiveness and generalization ability of the newly proposed SMCNet method.

Local Spatial–Spectral Information-Integrated Semisupervised Two-Stream Network for Hyperspectral Anomaly Detection

Spatial Anomaly Detection in Hyperspectral Imaging Using Optical Neural Networks

Hyperspectral Anomaly Detection Via Locally Enhanced Low-Rank Prior

Enhanced Tensor Low-Rank Representation Learning for Hyperspectral Anomaly Detection

Hyperspectral Band Selection for Spectral–Spatial Anomaly Detection

Spectral–Spatial Feature Extraction for Hyperspectral Anomaly Detection

Local Spatial Constraint and Total Variation for Hyperspectral Anomaly Detection

A hyperspectral anomaly detection framework based on segmentation and convolutional neural network algorithms

Multiple Spatial–Spectral Features Aggregated Neural Network for Hyperspectral Change Detection

MSNet: Self-Supervised Multiscale Network With Enhanced Separation Training for Hyperspectral Anomaly Detection

Spatial-Spectral Extraction for Hyperspectral Anomaly Detection

Exploiting the Sparse Characteristics in Probabilistic Feature Space for Hyperspectral Anomaly Detection

A Spatial-Spectral-Temporal Attention Method for Hyperspectral Image Change Detection.

Hyperspectral Anomaly Detection with Data Sphering and Unsupervised Target Detection.

Robust Tensor Low-Rank Sparse Representation with Saliency Prior for Hyperspectral Anomaly Detection

Spectral-Spatial-Temporal Attention Network for Hyperspectral Tracking.

Coarse-To-Fine Multiview Anomaly Coupling Network for Hyperspectral Anomaly Detection

Hyperspectral deep convolution anomaly detection based on weight adjustment strategy

Hyperspectral Anomaly Detection based on Machine Learning: An Overview

SMCNet: Sparse-Inspired Masked Convolutional Network for Hyperspectral Anomaly Detection

Self-Adaptive Low-Rank and Sparse Decomposition for Hyperspectral Anomaly Detection