Search

Your search keyword '"Liu, Haijun"' showing total 641 results
Start Over Author "Liu, Haijun" Publication Year Range Last 3 years
641 results on '"Liu, Haijun"'

8. BiGSeT: Binary Mask-Guided Separation Training for DNN-based Hyperspectral Anomaly Detection

Author

Liu, Haijun, Su, Xi, Shen, Xiangfei, Chen, Lihui, and Zhou, Xichuan

Subjects
Electrical Engineering and Systems Science - Image and Video Processing
Abstract

Hyperspectral anomaly detection (HAD) aims to recognize a minority of anomalies that are spectrally different from their surrounding background without prior knowledge. Deep neural networks (DNNs), including autoencoders (AEs), convolutional neural networks (CNNs) and vision transformers (ViTs), have shown remarkable performance in this field due to their powerful ability to model the complicated background. However, for reconstruction tasks, DNNs tend to incorporate both background and anomalies into the estimated background, which is referred to as the identical mapping problem (IMP) and leads to significantly decreased performance. To address this limitation, we propose a model-independent binary mask-guided separation training strategy for DNNs, named BiGSeT. Our method introduces a separation training loss based on a latent binary mask to separately constrain the background and anomalies in the estimated image. The background is preserved, while the potential anomalies are suppressed by using an efficient second-order Laplacian of Gaussian (LoG) operator, generating a pure background estimate. In order to maintain separability during training, we periodically update the mask using a robust proportion threshold estimated before the training. In our experiments, We adopt a vanilla AE as the network to validate our training strategy on several real-world datasets. Our results show superior performance compared to some state-of-the-art methods. Specifically, we achieved a 90.67% AUC score on the HyMap Cooke City dataset. Additionally, we applied our training strategy to other deep network structures, achieving improved detection performance compared to their original versions, demonstrating its effective transferability. The code of our method will be available at https://github.com/enter-i-username/BiGSeT., Comment: 13 pages, 13 figures, submitted to IEEE TRANSACTIONS ON IMAGE PROCESSING

Published
2023

23. Modified Stoney formula for obtainment of stress within thin films on large deformed isotropic circular plates

Author

Liu, Haijun, Dai, Minghui, Tian, Xiaoqing, Chen, Shan, Dong, Fangfang, and Lu, Lei

Subjects
Physics - Applied Physics
Abstract

Stoney formula is widely used to obtain the residual stress in the films on isotropic circular plates. However, significant errors would be introduced for large deformations because of the assumption of small deformation in the Stoney formula. In this study, a modified Stoney formula was proposed to extend its scope of application to the nonlinear domain. One-phase exponential decay function with coefficient p was used to relate the curvature of the substrate to the stress in the film. The coefficient p could be expressed as a function of the thickness, diameter, Young's modulus, and Poisson's ratio of the circular plate. The linear fitting technique was applied to ascertain the relationship between the coefficient p and the aforementioned parameters. The simple yet accurate modified Stoney formula could be used to calculate the residual stress in the film directly from the measured curvature of isotropic circular plates with various dimensions and materials.

Published
2021
Full Text
View/download PDF

28. The Impact of the Big Ideology Program on the Construction of Academic Ethos in Colleges and Universities in the Context of Three-Whole Parenting

Author

Liu Haijun and Tao Bingbing

Subjects
multiple linear regression, outliers, fitting, correlation coefficients, school spirituality, 01a74, Mathematics, QA1-939
Abstract

This paper uses multiple linear regression model to study the correlation between the ideological and political courses and the construction of university academic style, explores the calculation process of multiple linear regression, studies the impact analysis of linear regression model and the diagnosis of abnormal points, and analyzes the underfitting and overfitting respectively. The linear regression analysis model was used to study the differences in the academic style level of college students of different grades and genders, the correlation analysis of various dimensions of academic style and ideological and political level was carried out, and the linear fit of ideological and political curriculum education and academic style was analyzed by MATLAB to infer the significance level of ideological and political education and academic style. The results show that the correlation coefficient between ideology and politics and the dimension of learning purpose is 0.053, P=0.003

Published
2024
Full Text
View/download PDF

29. Analysis on the Modeling of Rockbusrt Prediction in Deep Tunnels Based on Machine Learning

Author

Zhang, Yong, Liu, Haijun, Ma, Yueyuan, Chen, Xueqi, Li, Weinan, Huang, Zheng, di Prisco, Marco, Series Editor, Chen, Sheng-Hong, Series Editor, Vayas, Ioannis, Series Editor, Kumar Shukla, Sanjay, Series Editor, Sharma, Anuj, Series Editor, Kumar, Nagesh, Series Editor, Wang, Chien Ming, Series Editor, Wang, Shuren, editor, Li, Jingan, editor, Hu, Kui, editor, and Bao, Xingxian, editor

Published
2023
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results