Search

Your search keyword '"Cui, Zhiying"' showing total 147 results
Start Over Author "Cui, Zhiying"
147 results on '"Cui, Zhiying"'

1. Decay Pruning Method: Smooth Pruning With a Self-Rectifying Procedure

Author

Yang, Minghao, Gao, Linlin, Li, Pengyuan, Li, Wenbo, Dong, Yihong, and Cui, Zhiying

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

Current structured pruning methods often result in considerable accuracy drops due to abrupt network changes and loss of information from pruned structures. To address these issues, we introduce the Decay Pruning Method (DPM), a novel smooth pruning approach with a self-rectifying mechanism. DPM consists of two key components: (i) Smooth Pruning: It converts conventional single-step pruning into multi-step smooth pruning, gradually reducing redundant structures to zero over N steps with ongoing optimization. (ii) Self-Rectifying: This procedure further enhances the aforementioned process by rectifying sub-optimal pruning based on gradient information. Our approach demonstrates strong generalizability and can be easily integrated with various existing pruning methods. We validate the effectiveness of DPM by integrating it with three popular pruning methods: OTOv2, Depgraph, and Gate Decorator. Experimental results show consistent improvements in performance compared to the original pruning methods, along with further reductions of FLOPs in most scenarios.

Published
2024

4. Continual Learning with Bayesian Model based on a Fixed Pre-trained Feature Extractor

Author

Yang, Yang, Cui, Zhiying, Xu, Junjie, Zhong, Changhong, Zheng, Wei-Shi, and Wang, Ruixuan

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

Deep learning has shown its human-level performance in various applications. However, current deep learning models are characterised by catastrophic forgetting of old knowledge when learning new classes. This poses a challenge particularly in intelligent diagnosis systems where initially only training data of a limited number of diseases are available. In this case, updating the intelligent system with data of new diseases would inevitably downgrade its performance on previously learned diseases. Inspired by the process of learning new knowledge in human brains, we propose a Bayesian generative model for continual learning built on a fixed pre-trained feature extractor. In this model, knowledge of each old class can be compactly represented by a collection of statistical distributions, e.g. with Gaussian mixture models, and naturally kept from forgetting in continual learning over time. Unlike existing class-incremental learning methods, the proposed approach is not sensitive to the continual learning process and can be additionally well applied to the data-incremental learning scenario. Experiments on multiple medical and natural image classification tasks showed that the proposed approach outperforms state-of-the-art approaches which even keep some images of old classes during continual learning of new classes.

Published
2022

5. Discriminative Distillation to Reduce Class Confusion in Continual Learning

Author

Zhong, Changhong, Cui, Zhiying, Wang, Ruixuan, and Zheng, Wei-Shi

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Successful continual learning of new knowledge would enable intelligent systems to recognize more and more classes of objects. However, current intelligent systems often fail to correctly recognize previously learned classes of objects when updated to learn new classes. It is widely believed that such downgraded performance is solely due to the catastrophic forgetting of previously learned knowledge. In this study, we argue that the class confusion phenomena may also play a role in downgrading the classification performance during continual learning, i.e., the high similarity between new classes and any previously learned classes would also cause the classifier to make mistakes in recognizing these old classes, even if the knowledge of these old classes is not forgotten. To alleviate the class confusion issue, we propose a discriminative distillation strategy to help the classify well learn the discriminative features between confusing classes during continual learning. Experiments on multiple natural image classification tasks support that the proposed distillation strategy, when combined with existing methods, is effective in further improving continual learning., Comment: arXiv admin note: text overlap with arXiv:2104.13614

Published
2021

8. Research on Early Warning Model of Wushu Event Broadcasting Right Operation Risk Based on Big Data XGBoost Algorithm

Author

Li Xing, Ma Ying, Cui Zhiying, and Cui Yongxia

Subjects
xgboost, gbrt, race broadcasting, risk warning, 68q05, Mathematics, QA1-939
Abstract

Under the background of the development of new media technology, the attention of wushu events in the society is gradually increasing, which makes the competition in the event broadcasting market more and more intense. This paper focuses on the problem of predicting the operational risk of wushu event broadcasting rights, based on the GBRT algorithm, innovatively improves the traditional loss function, introduces the regular term, and proposes the application of XGBoost algorithm in the operational risk prediction of wushu event broadcasting rights. The improved algorithm divides the operational risk of broadcasting rights into two main levels, covering three primary and 10 secondary indicators. In this study, the XGBoost algorithm is applied in the early warning of informing proper operation risk, which is classified into two main levels, covering 3 primary and 10 secondary indicators. The article also conducts an in-depth experimental analysis of the risk of overpremium of the event rights and the risk of matching the audience’s demand. In addition, according to the results of audience analysis, men have become the primary audience of wushu events, with a frequency of up to 401 times. Based on the XGBoost algorithm, the wushu event broadcasting right operation risk warning system can effectively predict and help the event broadcasting platform to avoid the potential operation risk, which provides valuable data support for the market decision-making.

Published
2024
Full Text
View/download PDF

10. Discriminative Distillation to Reduce Class Confusion in Continual Learning

Author

Zhong, Changhong, Cui, Zhiying, Zheng, Wei-Shi, Liu, Hongmei, Wang, Ruixuan, 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, Yu, Shiqi, editor, Zhang, Zhaoxiang, editor, Yuen, Pong C., editor, Han, Junwei, editor, Tan, Tieniu, editor, Guo, Yike, editor, Lai, Jianhuang, editor, and Zhang, Jianguo, editor

Published
2022
Full Text
View/download PDF

12. Positioning Error Modeling with Signal Burst of LEO Constellation

Author

Cui, Zhiying, Yue, Fuzhan, Tian, Run, Zhang, Shuangna, Dong, Qijia, Akan, Ozgur, Editorial Board Member, Bellavista, Paolo, Editorial Board Member, Cao, Jiannong, Editorial Board Member, Coulson, Geoffrey, Editorial Board Member, Dressler, Falko, Editorial Board Member, Ferrari, Domenico, Editorial Board Member, Gerla, Mario, Editorial Board Member, Kobayashi, Hisashi, Editorial Board Member, Palazzo, Sergio, Editorial Board Member, Sahni, Sartaj, Editorial Board Member, Shen, Xuemin (Sherman), Editorial Board Member, Stan, Mircea, Editorial Board Member, Jia, Xiaohua, Editorial Board Member, Zomaya, Albert Y., Editorial Board Member, Wang, Xianbin, editor, Wong, Kai-Kit, editor, Chen, Shanji, editor, and Liu, Mingqian, editor

Published
2021
Full Text
View/download PDF

13. Continual Learning with Bayesian Model Based on a Fixed Pre-trained Feature Extractor

Author

Yang, Yang, Cui, Zhiying, Xu, Junjie, Zhong, Changhong, Wang, Ruixuan, Zheng, Wei-Shi, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Woeginger, Gerhard, Editorial Board Member, Yung, Moti, Editorial Board Member, de Bruijne, Marleen, editor, Cattin, Philippe C., editor, Cotin, Stéphane, editor, Padoy, Nicolas, editor, Speidel, Stefanie, editor, Zheng, Yefeng, editor, and Essert, Caroline, editor

Published
2021
Full Text
View/download PDF

14. Compact Chromatic Confocal Lens with Large Measurement Range.

Author

He, Ning, Hu, Huiqin, Cui, Zhiying, Xu, Xinjun, Zhou, Dakai, Chen, Yunbo, Gong, Puyin, Chen, Youhua, and Kuang, Cuifang

Abstract

Spectral confocal sensors are effective for measuring displacements. The core of the spectral confocal measurement system is a dispersive objective lens that uses optical dispersion to establish a one-to-one correspondence between the focusing position and wavelength, achieving high-resolution measurements in the longitudinal direction. Despite significant progress in dispersive objective lenses for spectral confocal sensor systems, challenges such as a limited dispersion range, high cost, and insufficient measurement accuracy persist. To expand the measurement range and improve the accuracy of the spectral confocal sensor, we designed a compact, long-axial dispersion objective lens. This lens has a simple structure that requires only six lens elements, two of which form cemented doublets. The system length is 58 mm, with a working distance of 46 ± 6 mm and a dispersion range of 12 mm within the wavelength range of 450–656 nm. The lens has an object-side numerical aperture (NA) of 0.22 and an image-side NA between 0.198 and 0.24, ensuring high light energy utilization. Finally, a spectral confocal measurement system was constructed based on the designed dispersive objective lens, and performance evaluation tests were conducted. The test results showed that the system achieved a resolution of 0.15 μm and a maximum linear error of ±0.7 μm, demonstrating high-precision measurement capabilities. The proposed lens design enables the development of more portable and cost-effective spectral confocal sensors. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

19. Ensemble Transductive Learning for Skin Lesion Segmentation

Author

Cui, Zhiying, Wu, Longshi, Wang, Ruixuan, Zheng, Wei-Shi, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Woeginger, Gerhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Lin, Zhouchen, editor, Wang, Liang, editor, Yang, Jian, editor, Shi, Guangming, editor, Tan, Tieniu, editor, Zheng, Nanning, editor, Chen, Xilin, editor, and Zhang, Yanning, editor

Published
2019
Full Text
View/download PDF

31. Oral hygiene behaviours and their associated factors in Chinese adolescents: Evidence from the National Oral Health Survey

Author

Cui, Zhiying, primary, Wang, Wenhui, additional, Wang, Xing, additional, Feng, Xiping, additional, Tai, Baojun, additional, Hu, Deyu, additional, Lin, Huancai, additional, Wang, Bo, additional, Wang, Chunxiao, additional, Zheng, Shuguo, additional, Liu, Xuenan, additional, Rong, Wensheng, additional, Wang, Weijian, additional, and Si, Yan, additional

Published
2023
Full Text
View/download PDF

35. Oral hygiene behaviours and their associated factors in Chinese adolescents: Evidence from the National Oral Health Survey

Author

Wang, Wenhui, primary, Cui, Zhiying, additional, Du, Shuo, additional, Yu, Yizhen, additional, Xu, Kaige, additional, Wang, Mingyue, additional, Wang, Xing, additional, Feng, Xiping, additional, Tai, Baojun, additional, Hu, Deyu, additional, Lin, Huancai, additional, Wang, Bo, additional, Wang, Chunxiao, additional, Zheng, Shuguo, additional, Liu, Xuenan, additional, Rong, Wensheng, additional, Wang, Weijian, additional, and Si, Yan, additional

Published
2022
Full Text
View/download PDF

38. A Mitochondria-Targeted Phenylbutyric Acid Prodrug Confers Drastically Improved Anticancer Activities

Author

Huang, Ding, primary, Liu, Qingwang, additional, Zhang, Maojie, additional, Guo, Yizhen, additional, Cui, Zhiying, additional, Li, Tao, additional, Luo, Dong, additional, Xu, Biao, additional, Huang, Chao, additional, Guo, Jian, additional, Tam, Kin Yip, additional, Zhang, Min, additional, Zhang, Shao-Lin, additional, and He, Yun, additional

Published
2022
Full Text
View/download PDF

49. Proteomic Characterization of PAMs with PRRSV-ADE Infection.

Author

Xu, Pengli, Li, Wen, Zhao, Shijie, Cui, Zhiying, Chen, Yu, Zhang, Yi-na, Chen, Jing, and Xia, Pingan

Subjects
PROTEOMICS, ALVEOLAR macrophages, INFECTION, VIRAL replication, RIBOSOMES
Abstract

The antibody-dependent enhancement (ADE) effect of a PRRSV infection is that the preexisting sub- or non-neutralizing antibodies specific against PRRSV can facilitate the virus entry and replication, and it is likely to be a great obstacle for the selection of immune strategies and the development of high-efficiency PRRSV vaccines. However, the proteomic characterization of primary alveolar macrophages (PAMs) with a PRRSV-ADE infection has not yet been investigated so far. Therefore, we performed a tandem mass tag (TMT)-based quantitative proteomic analysis of PAMs with a PRRSV-ADE infection in this study. The results showed that a total of 3935 differentially expressed proteins (DEPs) were identified in the PAMs infected with PRRSV-ADE, including 2004 up-regulated proteins and 1931 down-regulated proteins. Further, the bioinformatics analysis for these DEPs revealed that a PRRSV-ADE infection might disturb the functions of ribosome, proteasome and mitochondria. Interestingly, we also found that the expression of the key molecules in the innate immune pathways and antiviral proteins were significantly down-regulated during a PRRSV-ADE infection. This study was the first attempt to analyze the proteomic characterization of PAMs with a PRRSV-ADE infection in vitro. Additionally, the findings will provide valuable information for a better understanding of the mechanism of virus–antibody–host interactions during a PRRSV-ADE infection. [ABSTRACT FROM AUTHOR]

Published
2023
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results