Search

Your search keyword '"Wang, Chenyang"' showing total 2,400 results
Start Over Author "Wang, Chenyang"
2,400 results on '"Wang, Chenyang"'

62. From Less to More: Spectral Splitting and Aggregation Network for Hyperspectral Face Super-Resolution

Author

Jiang, Junjun, Wang, Chenyang, Liu, Xianming, Jiang, Kui, and Ma, Jiayi

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

High-resolution (HR) hyperspectral face image plays an important role in face related computer vision tasks under uncontrolled conditions, such as low-light environment and spoofing attacks. However, the dense spectral bands of hyperspectral face images come at the cost of limited amount of photons reached a narrow spectral window on average, which greatly reduces the spatial resolution of hyperspectral face images. In this paper, we investigate how to adapt the deep learning techniques to hyperspectral face image super-resolution (HFSR), especially when the training samples are very limited. Benefiting from the amount of spectral bands, in which each band can be seen as an image, we present a spectral splitting and aggregation network (SSANet) for HFSR with limited training samples. In the shallow layers, we split the hyperspectral image into different spectral groups. Then, we gradually aggregate the neighbor bands at deeper layers to exploit spectral correlations. By this spectral splitting and aggregation strategy (SSAS), we can divide the original hyperspectral image into multiple samples (\emph{from less to more}) to support the efficient training of the network and effectively exploit the spectral correlations among spectrum. To cope with the challenge of small training sample size (S3) problem, we propose to expand the training samples by a self-representation model and symmetry-induced augmentation. Experiments show that SSANet can well model the joint correlations of spatial and spectral information. By expanding the training samples, SSANet can effectively alleviate the S3 problem., Comment: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

Published
2021

63. Efficient global sensitivity analysis method for passive systems under ocean conditions

Author

ZHANG Shiqi, PENG Minjun, XIA Genglei, WANG Chenyang, and SHANG He

Subjects
passive safety system, sensitivity analysis, low-rank approximations (lra), sobol method, ocean conditions, Nuclear engineering. Atomic power, TK9001-9401
Abstract

BackgroundPassive safety system reliability is generally evaluated through best estimation plus uncertainty (BEPU) analysis. An important step in the evaluation process is sensitivity analysis of parameters, which is used to identify key system parameters to reduce the complexity of the model. However, local sensitivity analysis methods based on linear or monotonic assumptions may yield incorrect sensitivity results for complex nuclear power systems. Meanwhile, applying the global sensitivity method is difficult in practical engineering because of its high calculation cost.PurposeThis study aims to develop an efficient and low-cost global sensitivity analysis method for passive systems under ocean conditions.MethodsFirstly, the low-rank approximation (LRA) method was employed to improve the Sobol method based on Monte Carlo simulation. The number of unknown coefficients was significantly reduced by using the multivariate-based tensor product. Then, the LRA coefficients were used to calculate the sensitivity index, and the validity of the proposed method was verified by addressing several sensitivity analysis benchmark questions. Finally, taking an integrated pressurized water reactor including the passive residual heat removal system as the object, a simulation program for thermal-hydraulics analysis under ocean conditions was developed. And its sensitivity analysis was conducted using the proposed method.ResultsThe results show that the proposed method can accurately identify system key parameters after only 200 simulation calculations taking about 55 min, and the sensitivity ranking results are consistent with those obtained by Sobol method after 1.0×105 simulation calculations taking about 19 d.ConclusionsThe efficient global sensitivity analysis method established in this study can provide effective guidance for reliability analysis and design optimization of passive systems.

Published
2024
Full Text
View/download PDF

65. Akkermansia muciniphila ameliorates colonic injury in mice with DSS-induced acute colitis by blocking macrophage pro-inflammatory phenotype switching via the HDAC5/DAB2 axis

Author

Miao, Yan, Wang, Mian, Sun, Hao, Zhang, Yujie, Zhou, Wei, Yang, Wanli, Duan, Lili, Niu, Liaoran, Li, Zhenshun, Chen, Junfeng, Li, Yiding, Fan, Aqiang, Xie, Qibin, Wei, Siyu, Bai, Han, Wang, Chenyang, Chen, Qian, Wang, Xiangjie, Li, Yunlong, Liu, Jinqiang, Han, Yu, Fan, Daiming, and Hong, Liu

Published
2024
Full Text
View/download PDF

74. Learning with Noisy Labels via Sparse Regularization

Author

Zhou, Xiong, Liu, Xianming, Wang, Chenyang, Zhai, Deming, Jiang, Junjun, and Ji, Xiangyang

Subjects
Computer Science - Machine Learning, Computer Science - Computer Vision and Pattern Recognition, Statistics - Machine Learning
Abstract

Learning with noisy labels is an important and challenging task for training accurate deep neural networks. Some commonly-used loss functions, such as Cross Entropy (CE), suffer from severe overfitting to noisy labels. Robust loss functions that satisfy the symmetric condition were tailored to remedy this problem, which however encounter the underfitting effect. In this paper, we theoretically prove that \textbf{any loss can be made robust to noisy labels} by restricting the network output to the set of permutations over a fixed vector. When the fixed vector is one-hot, we only need to constrain the output to be one-hot, which however produces zero gradients almost everywhere and thus makes gradient-based optimization difficult. In this work, we introduce the sparse regularization strategy to approximate the one-hot constraint, which is composed of network output sharpening operation that enforces the output distribution of a network to be sharp and the $\ell_p$-norm ($p\le 1$) regularization that promotes the network output to be sparse. This simple approach guarantees the robustness of arbitrary loss functions while not hindering the fitting ability. Experimental results demonstrate that our method can significantly improve the performance of commonly-used loss functions in the presence of noisy labels and class imbalance, and outperform the state-of-the-art methods. The code is available at https://github.com/hitcszx/lnl_sr., Comment: ICCV 2021

Published
2021

79. Definitive single fraction spine stereotactic radiosurgery for metastatic sarcoma: Simultaneous integrated boost is associated with high tumor control and low vertebral fracture risk

Author

Shanker, Mihir D., Cavazos, Adriana P., Li, Jing, Beckham, Thomas H., Yeboa, Debra N., Wang, Chenyang, McAleer, Mary Frances, Briere, Tina Marie, Amini, Behrang, Tatsui, Claudio E., North, Robert Y., Alvarez-Breckenridge, Christopher A., Cezayirli, Phillip Cem, Rhines, Laurence D., Ghia, Amol J., and Bishop, Andrew J.

Published
2024
Full Text
View/download PDF

89. PML: Progressive Margin Loss for Long-tailed Age Classification

Author

Deng, Zongyong, Liu, Hao, Wang, Yaoxing, Wang, Chenyang, Yu, Zekuan, and Sun, Xuehong

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

In this paper, we propose a progressive margin loss (PML) approach for unconstrained facial age classification. Conventional methods make strong assumption on that each class owns adequate instances to outline its data distribution, likely leading to bias prediction where the training samples are sparse across age classes. Instead, our PML aims to adaptively refine the age label pattern by enforcing a couple of margins, which fully takes in the in-between discrepancy of the intra-class variance, inter-class variance and class center. Our PML typically incorporates with the ordinal margin and the variational margin, simultaneously plugging in the globally-tuned deep neural network paradigm. More specifically, the ordinal margin learns to exploit the correlated relationship of the real-world age labels. Accordingly, the variational margin is leveraged to minimize the influence of head classes that misleads the prediction of tailed samples. Moreover, our optimization carefully seeks a series of indicator curricula to achieve robust and efficient model training. Extensive experimental results on three face aging datasets demonstrate that our PML achieves compelling performance compared to state of the arts. Code will be made publicly., Comment: Accepted at CVPR2021

Published
2021

90. Comparison of Response to Definitive Radiotherapy for Localized Prostate Cancer in Black and White Men

Author

Martin, Ting, Romero, Tahmineh, Nickols, Nicholas G, Rettig, Matthew B, Garraway, Isla P, Roach, Mack, Michalski, Jeff M, Pisansky, Thomas M, Lee, W Robert, Jones, Christopher U, Rosenthal, Seth A, Wang, Chenyang, Hartman, Holly, Nguyen, Paul L, Feng, Felix Y, Boutros, Paul C, Saigal, Christopher, Chamie, Karim, Jackson, William C, Morgan, Todd M, Mehra, Rohit, Salami, Simpa S, Vince, Randy, Schaeffer, Edward M, Mahal, Brandon A, Dess, Robert T, Steinberg, Michael L, Elashoff, David, Sandler, Howard M, Spratt, Daniel E, and Kishan, Amar U

Subjects
Clinical Trials and Supportive Activities, Cancer, Prostate Cancer, Urologic Diseases, Clinical Research, Good Health and Well Being, Black People, Humans, Male, Prostatic Neoplasms, Randomized Controlled Trials as Topic, Treatment Outcome, White People
Abstract

ImportanceBlack men have a 2-fold increased risk of dying from prostate cancer compared with White men. However, race-specific differences in response to initial treatment remain unknown.ObjectiveTo compare overall and treatment-specific outcomes of Black and White men with localized prostate cancer receiving definitive radiotherapy (RT).Data sourcesA systematic search was performed of relevant published randomized clinical trials conducted by the NRG Oncology/Radiation Therapy Oncology Group between January 1, 1990, and December 31, 2010. This meta-analysis was performed from July 1, 2019, to July 1, 2021.Study selectionRandomized clinical trials of definitive RT for patients with localized prostate cancer comprising a substantial number of Black men (self-identified race) enrolled that reported on treatment-specific and overall outcomes.Data extraction and synthesisIndividual patient data were obtained from 7 NRG Oncology/Radiation Therapy Oncology Group randomized clinical trials evaluating definitive RT with or without short- or long-term androgen deprivation therapy. Unadjusted Fine-Gray competing risk models, with death as a competing risk, were developed to evaluate the cumulative incidences of end points. Cox proportional hazards models were used to evaluate differences in all-cause mortality and the composite outcome of distant metastasis (DM) or death. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline was followed.Main outcomes and measuresSubdistribution hazard ratios (sHRs) of biochemical recurrence (BCR), DM, and prostate cancer-specific mortality (PCSM).ResultsA total of 8814 patients (1630 [18.5%] Black and 7184 [81.5%] White) were included; mean (SD) age was 69.1 (6.8) years. Median follow-up was 10.6 (IQR, 8.0-17.8) years for surviving patients. At enrollment, Black men were more likely to have high-risk disease features. However, even without adjustment, Black men were less likely to experience BCR (sHR, 0.88; 95% CI, 0.58-0.91), DM (sHR, 0.72; 95% CI, 0.58-0.91), or PCSM (sHR, 0.72; 95% CI, 0.54-0.97). No significant differences in all-cause mortality were identified (HR, 0.99; 95% CI, 0.92-1.07). Upon adjustment, Black race remained significantly associated with improved BCR (adjusted sHR, 0.79; 95% CI, 0.72-0.88; P

Published
2021

91. Design and Analysis of a High-performance Flexible Joint Actuator Based on the Peano-HASEL Actuator

Author

Sun, Wenjie, Liang, Huwei, Wang, Chenyang, Zhang, Fei, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Yang, Huayong, editor, Liu, Honghai, editor, Zou, Jun, editor, Yin, Zhouping, editor, Liu, Lianqing, editor, Yang, Geng, editor, Ouyang, Xiaoping, editor, and Wang, Zhiyong, editor

Published
2023
Full Text
View/download PDF

92. Quasi-Periodicity Detection via Repetition Invariance of Path Signatures

Author

Wang, Chenyang, Luo, Ling, Aickelin, Uwe, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Kashima, Hisashi, editor, Ide, Tsuyoshi, editor, and Peng, Wen-Chih, editor

Published
2023
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results