Category Ranking
98%
Total Visits
921
Avg Visit Duration
2 minutes
Citations
20
Article Abstract
Hyperspectral image (HSI) classification methods have made great progress in recent years. However, most of these methods are rooted in the closed-set assumption that the class distribution in the training and testing stages is consistent, which cannot handle the unknown class in open-world scenes. In this work, we propose a feature consistency-based prototype network (FCPN) for open-set HSI classification, which is composed of three steps. First, a three-layer convolutional network is designed to extract the discriminative features, where a contrastive clustering module is introduced to enhance the discrimination. Then, the extracted features are used to construct a scalable prototype set. Finally, a prototype-guided open-set module (POSM) is proposed to identify the known samples and unknown samples. Extensive experiments reveal that our method achieves remarkable classification performance over other state-of-the-art classification techniques.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1109/TNNLS.2022.3232225 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
SSC-PPI: A subspace structure consistency-based method for protein-protein interactions prediction.
IEEE Trans Comput Biol Bioinform
August 2025
Protein-protein interactions (PPIs) play an indispensable role in understanding disease-causing mechanisms, and the basic laws of food and drugs on life. Contemporary research on this issue, however, is incapable of guaranteeing structure consistency between extracted features and raw data, and fails to fully investigate the interconnection information of features. Thus, this paper proposes a subspace structure consistency-based method for protein-protein interactions prediction.
View Article and Find Full Text PDFThe demand for high-quality, lightweight infrared imaging systems is rapidly increasing. Single-lens computational imaging, combining single-lens with post-processing algorithms, offers a promising solution to miniaturize imaging systems while maintaining performance. However, these post-processing algorithms are typically highly complex, posing significant challenges for real-time reconstruction on a neural network processing unit (NPU) chip.
View Article and Find Full Text PDFDual-Student Adversarial Framework With Discriminator and Consistency-Driven Learning for Semi-Supervised Medical Image Segmentation.
IEEE J Biomed Health Inform
August 2025
Semi-supervised medical image segmentation is essential for alleviating the cost of manual annotation in clinical applications. However, existing methods often suffer from unreliable pseudo-labels and confirmation bias in consistency-based training, which can lead to unstable optimization and degraded performance. To address these issues, a novel method named dual-Student adversarial framework with discriminator and consistency-driven learning for semi-supervised medical image segmentation is proposed.
View Article and Find Full Text PDFA Multimodal Consistency-Based Self-Supervised Contrastive Learning Framework for Automated Sleep Staging in Patients With Disorders of Consciousness.
IEEE J Biomed Health Inform
February 2025
Article Synopsis
- Sleep staging is important for understanding sleep patterns, yet current deep learning methods face challenges like generalization issues and the lack of labeled data, particularly in patients with disorders of consciousness (DOC).
- The paper introduces MultiConsSleepNet, a network designed to extract universal features from EEGs and EOGs using both unimodal and multimodal approaches, which helps address the scarcity of labeled data through self-supervised learning.
- Experimental results show that MultiConsSleepNet performs exceptionally well in sleep staging across different datasets, demonstrating its ability to work effectively with limited labeled data and making it a promising tool for research involving DOC patients.
Unsupervised and Self-supervised Learning in Low-Dose Computed Tomography Denoising: Insights from Training Strategies.
J Imaging Inform Med
April 2025
School of Data Science and Artificial Intelligence, Wenzhou University of Technology, Ouhai District, Wenzhou, 325000, Zhejiang, China.
Article Synopsis
- X-ray low-dose computed tomography (LDCT) is recognized for reducing patient radiation exposure, but its images often suffer from noise that impacts diagnostic quality.
- Numerous LDCT denoising methods have been proposed, with deep learning techniques proving to be the most effective, although acquiring paired low-dose and normal-dose CT scans poses challenges for supervised training.
- This review categorizes unsupervised and self-supervised deep learning denoising methods into six groups and discusses their theoretical foundations, training strategies, strengths, and weaknesses, while also identifying open-source codes and future research directions.