Search

Your search keyword '"Zhang, Yunxuan"' showing total 140 results
Start Over Author "Zhang, Yunxuan"
140 results on '"Zhang, Yunxuan"'

1. Unified Pretraining Target Based Video-music Retrieval With Music Rhythm And Video Optical Flow Information

Author

Mao, Tianjun, Liu, Shansong, Zhang, Yunxuan, Li, Dian, and Shan, Ying

Subjects
Computer Science - Multimedia
Abstract

Background music (BGM) can enhance the video's emotion. However, selecting an appropriate BGM often requires domain knowledge. This has led to the development of video-music retrieval techniques. Most existing approaches utilize pretrained video/music feature extractors trained with different target sets to obtain average video/music-level embeddings. The drawbacks are two-fold. One is that different target sets for video/music pretraining may cause the generated embeddings difficult to match. The second is that the underlying temporal correlation between video and music is ignored. In this paper, our proposed approach leverages a unified target set to perform video/music pretraining and produces clip-level embeddings to preserve temporal information. The downstream cross-modal matching is based on the clip-level features with embedded music rhythm and optical flow information. Experiments demonstrate that our proposed method can achieve superior performance over the state-of-the-art methods by a significant margin.

Published
2023

9. TAD: Tracking-Aided Detection Siamese Network for Visual Drone Surveillance

Author

Zhang, Yunxuan, Angrisani, Leopoldo, Series Editor, Arteaga, Marco, Series Editor, Panigrahi, Bijaya Ketan, Series Editor, Chakraborty, Samarjit, Series Editor, Chen, Jiming, Series Editor, Chen, Shanben, Series Editor, Chen, Tan Kay, Series Editor, Dillmann, Rüdiger, Series Editor, Duan, Haibin, Series Editor, Ferrari, Gianluigi, Series Editor, Ferre, Manuel, Series Editor, Hirche, Sandra, Series Editor, Jabbari, Faryar, Series Editor, Jia, Limin, Series Editor, Kacprzyk, Janusz, Series Editor, Khamis, Alaa, Series Editor, Kroeger, Torsten, Series Editor, Li, Yong, Series Editor, Liang, Qilian, Series Editor, Martín, Ferran, Series Editor, Ming, Tan Cher, Series Editor, Minker, Wolfgang, Series Editor, Misra, Pradeep, Series Editor, Möller, Sebastian, Series Editor, Mukhopadhyay, Subhas, Series Editor, Ning, Cun-Zheng, Series Editor, Nishida, Toyoaki, Series Editor, Oneto, Luca, Series Editor, Pascucci, Federica, Series Editor, Qin, Yong, Series Editor, Seng, Gan Woon, Series Editor, Speidel, Joachim, Series Editor, Veiga, Germano, Series Editor, Wu, Haitao, Series Editor, Zamboni, Walter, Series Editor, Zhang, Junjie James, Series Editor, Fu, Wenxing, editor, Gu, Mancang, editor, and Niu, Yifeng, editor

Published
2023
Full Text
View/download PDF

12. One-shot Face Reenactment

Author

Zhang, Yunxuan, Zhang, Siwei, He, Yue, Li, Cheng, Loy, Chen Change, and Liu, Ziwei

Subjects
Computer Science - Computer Vision and Pattern Recognition, Electrical Engineering and Systems Science - Image and Video Processing
Abstract

To enable realistic shape (e.g. pose and expression) transfer, existing face reenactment methods rely on a set of target faces for learning subject-specific traits. However, in real-world scenario end-users often only have one target face at hand, rendering existing methods inapplicable. In this work, we bridge this gap by proposing a novel one-shot face reenactment learning framework. Our key insight is that the one-shot learner should be able to disentangle and compose appearance and shape information for effective modeling. Specifically, the target face appearance and the source face shape are first projected into latent spaces with their corresponding encoders. Then these two latent spaces are associated by learning a shared decoder that aggregates multi-level features to produce the final reenactment results. To further improve the synthesizing quality on mustache and hair regions, we additionally propose FusionNet which combines the strengths of our learned decoder and the traditional warping method. Extensive experiments show that our one-shot face reenactment system achieves superior transfer fidelity as well as identity preserving capability than alternatives. More remarkably, our approach trained with only one target image per subject achieves competitive results to those using a set of target images, demonstrating the practical merit of this work. Code, models and an additional set of reenacted faces have been publicly released at the project page., Comment: To appear in BMVC 2019 as a spotlight presentation. Code and models are available at: https://github.com/bj80heyue/One_Shot_Face_Reenactment

Published
2019

13. CMR motion artifact correction using generative adversarial nets

Author

Zhang, Yunxuan, Zhang, Weiliang, Zhang, Qinyan, Yang, Jijiang, Chen, Xiuyu, and Zhao, Shihua

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

Cardiovascular Magnetic Resonance (CMR) plays an important role in the diagnoses and treatment of cardiovascular diseases while motion artifacts which are formed during the scanning process of CMR seriously affects doctors to find the exact focus. The current correction methods mainly focus on the K-space which is a grid of raw data obtained from the MR signal directly and then transfer to CMR image by inverse Fourier transform. They are neither effective nor efficient and can not be utilized in clinic. In this paper, we propose a novel approach for CMR motion artifact correction using deep learning. Specially, we use deep residual network (ResNet) as net framework and train our model in adversarial manner. Our approach is motivated by the connection between image motion blur and CMR motion artifact, so we can transfer methods from motion-deblur where deep learning has made great progress to CMR motion-correction successfully. To evaluate motion artifact correction methods, we propose a novel algorithm on how edge detection results are improved by deblurred algorithm. Boosted by deep learning and adversarial training algorithm, our model is trainable in an end-to-end manner, can be tested in real-time and achieves the state-of-art results for CMR correction., Comment: 6 pages, 7 figures

Published
2019

14. Stress-Strain State of the Floating Bollard’s Structure for a Shipping Gateway

Author

Sydorenko, Ihor, Tonkonogyi, Vladimir, Semenyuk, Vladimir, Lingur, Valeriy, Zhang, Yunxuan, Cavas-Martínez, Francisco, Editorial Board Member, Chaari, Fakher, Series Editor, di Mare, Francesca, Editorial Board Member, Gherardini, Francesco, Series Editor, Haddar, Mohamed, Editorial Board Member, Ivanov, Vitalii, Series Editor, Kwon, Young W., Editorial Board Member, Trojanowska, Justyna, Editorial Board Member, Pavlenko, Ivan, editor, Liaposhchenko, Oleksandr, editor, Machado, José, editor, and Edl, Milan, editor

Published
2022
Full Text
View/download PDF

23. ReenactGAN: Learning to Reenact Faces via Boundary Transfer

Author

Wu, Wayne, Zhang, Yunxuan, Li, Cheng, Qian, Chen, and Loy, Chen Change

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Artificial Intelligence, Computer Science - Graphics
Abstract

We present a novel learning-based framework for face reenactment. The proposed method, known as ReenactGAN, is capable of transferring facial movements and expressions from monocular video input of an arbitrary person to a target person. Instead of performing a direct transfer in the pixel space, which could result in structural artifacts, we first map the source face onto a boundary latent space. A transformer is subsequently used to adapt the boundary of source face to the boundary of target face. Finally, a target-specific decoder is used to generate the reenacted target face. Thanks to the effective and reliable boundary-based transfer, our method can perform photo-realistic face reenactment. In addition, ReenactGAN is appealing in that the whole reenactment process is purely feed-forward, and thus the reenactment process can run in real-time (30 FPS on one GTX 1080 GPU). Dataset and model will be publicly available at https://wywu.github.io/projects/ReenactGAN/ReenactGAN.html, Comment: Accepted to ECCV 2018. Project page: https://wywu.github.io/projects/ReenactGAN/ReenactGAN.html

Published
2018

24. Quantifying Facial Age by Posterior of Age Comparisons

Author

Zhang, Yunxuan, Liu, Li, Li, Cheng, and Loy, Chen change

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

We introduce a novel approach for annotating large quantity of in-the-wild facial images with high-quality posterior age distribution as labels. Each posterior provides a probability distribution of estimated ages for a face. Our approach is motivated by observations that it is easier to distinguish who is the older of two people than to determine the person's actual age. Given a reference database with samples of known ages and a dataset to label, we can transfer reliable annotations from the former to the latter via human-in-the-loop comparisons. We show an effective way to transform such comparisons to posterior via fully-connected and SoftMax layers, so as to permit end-to-end training in a deep network. Thanks to the efficient and effective annotation approach, we collect a new large-scale facial age dataset, dubbed `MegaAge', which consists of 41,941 images. Data can be downloaded from our project page mmlab.ie.cuhk.edu.hk/projects/MegaAge and github.com/zyx2012/Age_estimation_BMVC2017. With the dataset, we train a network that jointly performs ordinal hyperplane classification and posterior distribution learning. Our approach achieves state-of-the-art results on popular benchmarks such as MORPH2, Adience, and the newly proposed MegaAge., Comment: To appear on BMVC 2017 (oral) revised version

Published
2017

25. Propensity score weighted multi‐source exchangeability models for incorporating external control data in randomized clinical trials.

Author

Wei, Wei, Zhang, Yunxuan, and Roychoudhury, Satrajit

Subjects
*CLINICAL trials, *RARE diseases, *DRUG development, *TREATMENT effectiveness
Abstract

Among clinical trialists, there has been a growing interest in using external data to improve decision‐making and accelerate drug development in randomized clinical trials (RCTs). Here we propose a novel approach that combines the propensity score weighting (PW) and the multi‐source exchangeability modelling (MEM) approaches to augment the control arm of a RCT in the rare disease setting. First, propensity score weighting is used to construct weighted external controls that have similar observed pre‐treatment characteristics as the current trial population. Next, the MEM approach evaluates the similarity in outcome distributions between the weighted external controls and the concurrent control arm. The amount of external data we borrow is determined by the similarities in pretreatment characteristics and outcome distributions. The proposed approach can be applied to binary, continuous and count data. We evaluate the performance of the proposed PW‐MEM method and several competing approaches based on simulation and re‐sampling studies. Our results show that the PW‐MEM approach improves the precision of treatment effect estimates while reducing the biases associated with borrowing data from external sources. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

33. Zearalenone Induces Blood-Testis Barrier Damage through Endoplasmic Reticulum Stress-Mediated Paraptosis of Sertoli Cells in Goats.

Author

Liu, Tengfei, Liu, Gengchen, Xu, Yinghuan, Huang, Yuqi, Zhang, Yunxuan, Wu, Yongjie, and Xu, Yongping

Subjects
SERTOLI cells, MALE infertility, ENDOPLASMIC reticulum, GOATS, POISONS, SEMINIFEROUS tubules, SPERMATOZOA
Abstract

Zearalenone (ZEA) is present worldwide as a serious contaminant of food and feed and causes male reproductive toxicity. The implication of paraptosis, which is a nonclassical paradigm of cell death, is unclear in ZEA-induced male reproductive disorders. In this study, the toxic effects of ZEA on the blood-testis barrier (BTB) and the related mechanisms of paraptosis were detected in goats. ZEA exposure, in vivo, caused a significant decrease in spermatozoon quality, the destruction of seminiferous tubules, and damage to the BTB integrity. Furthermore, ZEA exposure to Sertoli cells (SCs) in vitro showed similar dysfunction in structure and barrier function. Importantly, the formation of massive cytoplasmic vacuoles in ZEA-treated SCs corresponded to the highly swollen and dilative endoplasmic reticulum (ER), and paraptosis inhibition significantly alleviated ZEA-induced SC death and vacuolization, which indicated the important contribution of paraptosis in ZEA-induced BTB damage. Meanwhile, the expression of ER stress marker proteins was increased after ZEA treatment but decreased under the inhibition of paraptosis. The vacuole formation and SC death, induced by ZEA, were remarkably blocked by ER stress inhibition. In conclusion, these results facilitate the exploration of the mechanisms of the SC paraptosis involved in ZEA-induced BTB damage in goats. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

37. Benchmarking of local genetic correlation estimation methods using summary statistics from genome-wide association studies.

Author

Zhang, Chi, Zhang, Yiliang, Zhang, Yunxuan, and Zhao, Hongyu

Subjects
GENOME-wide association studies, GENETIC correlations, GENETIC variation, LINKAGE disequilibrium
Abstract

Local genetic correlation evaluates the correlation of additive genetic effects between different traits across the same genetic variants at a genomic locus. It has been proven informative for understanding the genetic similarities of complex traits beyond that captured by global genetic correlation calculated across the whole genome. Several summary-statistics-based approaches have been developed for estimating local genetic correlation, including |$\rho$| -hess, SUPERGNOVA and LAVA. However, there has not been a comprehensive evaluation of these methods to offer practical guidelines on the choices of these methods. In this study, we conduct benchmark comparisons of the performance of these three methods through extensive simulation and real data analyses. We focus on two technical difficulties in estimating local genetic correlation: sample overlaps across traits and local linkage disequilibrium (LD) estimates when only the external reference panels are available. Our simulations suggest the likelihood of incorrectly identifying correlated regions and local correlation estimation accuracy are highly dependent on the estimation of the local LD matrix. These observations are corroborated by real data analyses of 31 complex traits. Overall, our findings illuminate the distinct results yielded by different methods applied in post-genome-wide association studies (post-GWAS) local correlation studies. We underscore the sensitivity of local genetic correlation estimates and inferences to the precision of local LD estimation. These observations accentuate the vital need for ongoing refinement in methodologies. [ABSTRACT FROM AUTHOR]

Published
2023
Full Text
View/download PDF

38. The upregulation of peripheral CD3-CD56+CD16+ natural killer cells correlates with Th1/Th2 imbalance in asthma patients during acute upper respiratory viral infections.

Author

Liu, Meixuan, Zhang, Yunxuan, Hu, Yunqian, Guo, Zhongliang, and Dong, Lin

Subjects
*RESPIRATORY infections, *ASTHMATICS, *KILLER cells, *CELL populations, *VIRUS diseases
Abstract

Purpose: The aim of this study is to clarify the changes of peripheral CD3−CD56+CD16+ NK cells and their correlation with Th1/Th2 immunity profiles in asthma during the phase of acute upper respiratory viral infections (AURVIs). Methods: Peripheral venous blood and induced sputum samples were collected from 56 mild asthma patients, 49 asthma patients with AURVIs and 50 healthy subjects. Peripheral CD3−CD56+CD16+ NK cells were monitored by flow cytometry during the course of acute viral infections. Meanwhile, the induced sputum Th2 cytokines IL-4 and IL-5, and Th1 cytokine IFN-γ were also detected by ELISA assay. Results: The asthmatics had lower levels of peripheral CD3−CD56+CD16+ NK cells populations as well as higher induced sputum cytokines (IL-4, IL-5 and IFN-γ) compared to healthy controls at baseline. Upon upper respiratory viral infections, peripheral CD3−CD56+CD16+ NK cells numbers in asthma patients sharply elevated on day 3 and slowly decreased by day 14, in accordance with induced sputum IFN-γ changes. IL-4 and IL-5 levels spiked much later (day 8) and lasted until day 14. Compared with asthma alone group, the IFN-γ/IL-4 and IFN-γ/IL-5 ratios of the asthma patients with AURVIs on day 1 were higher and peaked on day 3. The changes of peripheral CD3−CD56+CD16+ NK cells proportions positively correlated with the IFN-γ/IL-4 and IFN-γ/IL-5 ratios on day 1 to day 3 in asthma subsequent to upper respiratory viral infections. Conclusions: Our findings showed an imbalanced Th1/Th2 immunity in airways of asthma with acute upper respiratory viral infections. Upregulated peripheral CD3−CD56+CD16+ NK cells play a crucial role in biased Th1 immunity of airways in asthma during the acute phase of viral infections. The anti-viral Th1 immunity by targeting NK cells may be a possible therapeutic option for virus-induced asthma exacerbation. [ABSTRACT FROM AUTHOR]

Published
2023
Full Text
View/download PDF

41. "Best Employers": The Impacts of Employee Reviews and Employer Awards on Job Seekers' Application Intentions.

Author

Zhang, Yunxuan, Shum, Cass, and Belarmino, Amanda

Subjects
EMPLOYEE reviews, ELABORATION likelihood model, INDUSTRIAL relations, MARKET tightness, USER-generated content
Abstract

While hospitality researchers have examined the impacts of user-generated content on customers, research regarding the impacts of employee reviews on job seekers' application intentions is scarce. Yet, labor shortages in the hospitality industry have been amplified in recent years. The tight job market requires organizations to use aggressive and proactive recruitment strategies. As online employee reviews can attract both active and passive job seekers, organizations are increasingly advertising their jobs on these sites. This study draws on the elaboration likelihood model (ELM) and tests the boundary condition of work experience on the effects of overall star-ratings and employer awards on job seekers' application intention. Through an experimental survey, this study sought to fill the gap regarding the impacts of employee-generated star-ratings and employer awards on job seekers' application intentions. Both star-ratings and employer awards are positively related to organizational prestige. Hospitality work experience moderates the relationship between star-ratings and organizational prestige. The relationship is stronger for novice job seekers than for experienced job seekers. Organizational prestige, in turn, increases job seekers' application intentions. Our findings extend the recruitment literature and highlight the potential usage of ELM as an explorative framework in hospitality recruitment research. The study also provides suggestions for hospitality employers to attract job seekers. [ABSTRACT FROM AUTHOR]

Published
2023
Full Text
View/download PDF

48. Optimization of the precision gear grinding operation based on integrated information system

Author

Vasily P. Larshin, Olga B. Babiychuk, Oleksandr V. Lysyi, Serhii M. Verpivskyi, and Zhang Yunxuan

Abstract

In accordance with the principles of hierarchical management, a comprehensive two-level management system is presented for the development and manufacturing of products for the stages of pre-production (the upper level of the management hierarchy) and for the actual production stage (the lower level of the management hierarchy). At the stage of pre-production, the gear grinding operation design on the “MAAG” type machines was carried out. For this purpose, a technique for optimizing the gear grinding parameters for a two dish-wheel rolling scheme has been developed, a mathematical optimization model containing an objective function with restrictions imposed on it has been created. The objective function is the gear grinding machine time, which depends on the operation parameters (gear grinding stock allowance, cutting modes, grinding wheel specification, part material) and the design features of the gears being ground (module, diameter, number of teeth, radius of curvature of the involutes). The article shows that at the stage of pre-production, the gear grinding optimization is a method of operation design. At the stage of actual production, a closed-loop automatic control system with feedback on the deviation of the adjustable value (gear grinding power) automatically supports the numerical power values that were found at the operation design stage, taking into account ensuring defect-free high-performance gear grinding (minimum number of working strokes and maximum longitudinal feeds). At this stage, i.e. when a robust longitudinal feed automatic control system is operating, the optimization carried out at the previous stage (pre-production) sets the functioning algorithm for the adaptive system with corresponding control algorithm. Thus, at the production stage (when the gear grinding machine is running), the operation optimization is a control method. Therefore, it is shown that with two-level control, the gear grinding operation optimization performs a dual function. On the one hand, it is a design method (at the pre-production stage), and on the other – a management method (at the actual production stage). With this approach, i.e. with the integration of production and its preparation based on a single two-level management, the efficiency of a single integrated design and production automation system is significantly higher due to general (unified) optimization, rather than partial one.

Published
2021

Catalog

Books, media, physical & digital resources
See catalog results