Bing Tu,Tao Zhou,Bo Liu,Yan He,Jun Li,Antonio Plaza
{"title":"高光谱图像异常检测中的多尺度自编码器抑制策略。","authors":"Bing Tu,Tao Zhou,Bo Liu,Yan He,Jun Li,Antonio Plaza","doi":"10.1109/tip.2025.3595408","DOIUrl":null,"url":null,"abstract":"Autoencoders (AEs) have received extensive attention in hyperspectral anomaly detection (HAD) due to their capability to separate the background from the anomaly based on the reconstruction error. However, the existing AE methods routinely fail to adequately exploit spatial information and may precisely reconstruct anomalies, thereby affecting the detection accuracy. To address these issues, this study proposes a novel Multi-scale Autoencoder Suppression Strategy (MASS). The underlying principle of MASS is to prioritize the reconstruction of background information over anomalies. In the encoding stage, the Local Feature Extractor, which integrates Convolution and Omni-Dimensional Dynamic Convolution (ODConv), is combined with the Global Feature Extractor based on Transformer to effectively extract multi-scale features. Furthermore, a Self-Attention Suppression module (SAS) is devised to diminish the influence of anomalous pixels, enabling the network to focus more intently on the precise reconstruction of the background. During the process of network learning, a mask derived from the test outcomes of each iteration is integrated into the loss function computation, encompassing only the positions with low anomaly scores from the preceding detection round. Experiments on eight datasets demonstrate that the proposed method is significantly superior to several traditional methods and deep learning methods in terms of performance.","PeriodicalId":13217,"journal":{"name":"IEEE Transactions on Image Processing","volume":"20 1","pages":""},"PeriodicalIF":13.7000,"publicationDate":"2025-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Multi-scale Autoencoder Suppression Strategy for Hyperspectral Image Anomaly Detection.\",\"authors\":\"Bing Tu,Tao Zhou,Bo Liu,Yan He,Jun Li,Antonio Plaza\",\"doi\":\"10.1109/tip.2025.3595408\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Autoencoders (AEs) have received extensive attention in hyperspectral anomaly detection (HAD) due to their capability to separate the background from the anomaly based on the reconstruction error. However, the existing AE methods routinely fail to adequately exploit spatial information and may precisely reconstruct anomalies, thereby affecting the detection accuracy. To address these issues, this study proposes a novel Multi-scale Autoencoder Suppression Strategy (MASS). The underlying principle of MASS is to prioritize the reconstruction of background information over anomalies. In the encoding stage, the Local Feature Extractor, which integrates Convolution and Omni-Dimensional Dynamic Convolution (ODConv), is combined with the Global Feature Extractor based on Transformer to effectively extract multi-scale features. Furthermore, a Self-Attention Suppression module (SAS) is devised to diminish the influence of anomalous pixels, enabling the network to focus more intently on the precise reconstruction of the background. During the process of network learning, a mask derived from the test outcomes of each iteration is integrated into the loss function computation, encompassing only the positions with low anomaly scores from the preceding detection round. Experiments on eight datasets demonstrate that the proposed method is significantly superior to several traditional methods and deep learning methods in terms of performance.\",\"PeriodicalId\":13217,\"journal\":{\"name\":\"IEEE Transactions on Image Processing\",\"volume\":\"20 1\",\"pages\":\"\"},\"PeriodicalIF\":13.7000,\"publicationDate\":\"2025-08-08\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"IEEE Transactions on Image Processing\",\"FirstCategoryId\":\"94\",\"ListUrlMain\":\"https://doi.org/10.1109/tip.2025.3595408\",\"RegionNum\":1,\"RegionCategory\":\"计算机科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE Transactions on Image Processing","FirstCategoryId":"94","ListUrlMain":"https://doi.org/10.1109/tip.2025.3595408","RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE","Score":null,"Total":0}
Multi-scale Autoencoder Suppression Strategy for Hyperspectral Image Anomaly Detection.
Autoencoders (AEs) have received extensive attention in hyperspectral anomaly detection (HAD) due to their capability to separate the background from the anomaly based on the reconstruction error. However, the existing AE methods routinely fail to adequately exploit spatial information and may precisely reconstruct anomalies, thereby affecting the detection accuracy. To address these issues, this study proposes a novel Multi-scale Autoencoder Suppression Strategy (MASS). The underlying principle of MASS is to prioritize the reconstruction of background information over anomalies. In the encoding stage, the Local Feature Extractor, which integrates Convolution and Omni-Dimensional Dynamic Convolution (ODConv), is combined with the Global Feature Extractor based on Transformer to effectively extract multi-scale features. Furthermore, a Self-Attention Suppression module (SAS) is devised to diminish the influence of anomalous pixels, enabling the network to focus more intently on the precise reconstruction of the background. During the process of network learning, a mask derived from the test outcomes of each iteration is integrated into the loss function computation, encompassing only the positions with low anomaly scores from the preceding detection round. Experiments on eight datasets demonstrate that the proposed method is significantly superior to several traditional methods and deep learning methods in terms of performance.
期刊介绍:
The IEEE Transactions on Image Processing delves into groundbreaking theories, algorithms, and structures concerning the generation, acquisition, manipulation, transmission, scrutiny, and presentation of images, video, and multidimensional signals across diverse applications. Topics span mathematical, statistical, and perceptual aspects, encompassing modeling, representation, formation, coding, filtering, enhancement, restoration, rendering, halftoning, search, and analysis of images, video, and multidimensional signals. Pertinent applications range from image and video communications to electronic imaging, biomedical imaging, image and video systems, and remote sensing.