Filters








3,465 Hits in 2.4 sec

Collaborative Unsupervised Visual Representation Learning from Decentralized Data [article]

Weiming Zhuang, Xin Gan, Yonggang Wen, Shuai Zhang, Shuai Yi
2021 arXiv   pre-print
Unsupervised representation learning has achieved outstanding performances using centralized data available on the Internet. However, the increasing awareness of privacy protection limits sharing of decentralized unlabeled image data that grows explosively in multiple parties (e.g., mobile phones and cameras). As such, a natural problem is how to leverage these data to learn visual representations for downstream tasks while preserving data privacy. To address this problem, we propose a novel
more » ... erated unsupervised learning framework, FedU. In this framework, each party trains models from unlabeled data independently using contrastive learning with an online network and a target network. Then, a central server aggregates trained models and updates clients' models with the aggregated model. It preserves data privacy as each party only has access to its raw data. Decentralized data among multiple parties are normally non-independent and identically distributed (non-IID), leading to performance degradation. To tackle this challenge, we propose two simple but effective methods: 1) We design the communication protocol to upload only the encoders of online networks for server aggregation and update them with the aggregated encoder; 2) We introduce a new module to dynamically decide how to update predictors based on the divergence caused by non-IID. The predictor is the other component of the online network. Extensive experiments and ablations demonstrate the effectiveness and significance of FedU. It outperforms training with only one party by over 5% and other methods by over 14% in linear and semi-supervised evaluation on non-IID data.
arXiv:2108.06492v1 fatcat:hkunizhgongzzgg26xu55kgesi

Federated Unsupervised Domain Adaptation for Face Recognition [article]

Weiming Zhuang, Xin Gan, Yonggang Wen, Xuesen Zhang, Shuai Zhang, Shuai Yi
2022 arXiv   pre-print
Given labeled data in a source domain, unsupervised domain adaptation has been widely adopted to generalize models for unlabeled data in a target domain, whose data distributions are different. However, existing works are inapplicable to face recognition under privacy constraints because they require sharing of sensitive face images between domains. To address this problem, we propose federated unsupervised domain adaptation for face recognition, FedFR. FedFR jointly optimizes clustering-based
more » ... omain adaptation and federated learning to elevate performance on the target domain. Specifically, for unlabeled data in the target domain, we enhance a clustering algorithm with distance constrain to improve the quality of predicted pseudo labels. Besides, we propose a new domain constraint loss (DCL) to regularize source domain training in federated learning. Extensive experiments on a newly constructed benchmark demonstrate that FedFR outperforms the baseline and classic methods on the target domain by 3% to 14% on different evaluation metrics.
arXiv:2204.04382v1 fatcat:kuf237uzs5hwnhcyxa2bcayeki

BST2 confers cisplatin resistance via NF-κB signaling in nasopharyngeal cancer

Chun-mei Kuang, Xiang Fu, Yi-jun Hua, Wen-di Shuai, Zhi-hua Ye, Yingchang Li, Qi-hua Peng, Yi-zhuo Li, Shuai Chen, Chao-nan Qian, Wenlin Huang, Ran-yi Liu
2017 Cell Death and Disease  
Concurrent/adjuvant cisplatin-based chemoradiotherapy is regarded as the standard of treatment for locoregionally advanced nasopharyngeal carcinoma (NPC). However, patients who do not respond to cisplatin suffer, rather than benefit, from chemotherapy treatment. The goal of this study was to identify molecules involved in cisplatin resistance and to clarify their molecular mechanisms, which would help in the discovery of potential therapeutic targets and in developing a personalized and precise
more » ... treatment approach for NPC patients. We previously generated a cisplatin-sensitive NPC cell line, S16, from CNE2 cells and found that eIF3a, ASNS and MMP19 are upregulated in S16 cells, which contributes to their cisplatin sensitivity. In this study, we found that BST2 is downregulated in cisplatin-sensitive S16 cells compared with CNE2 cells. Knockdown of BST2 in NPC cells sensitized their response to cisplatin and promoted cisplatin-induced apoptosis, whereas exogenous overexpression of BST2 increased their cisplatin resistance and inhibited cisplatin-induced apoptosis. Further investigation demonstrated that BST2-mediated cisplatin resistance depended on the activation of the NF-κB signaling pathway and consequent upregulation of anti-apoptotic genes, such as Bcl-X L and livin. Moreover, an analysis of clinical data revealed that a high BST2 level might serve as an independent indicator of poor prognosis in patients with locally advanced NPC treated with platinum-based chemoradiotherapy. These findings suggest that BST2 likely mediates platinum resistance in NPC, offering guidance for personalized and precise treatment strategies for patients with NPC.
doi:10.1038/cddis.2017.271 pmid:28617432 pmcid:PMC5520926 fatcat:tfctdk7nezccvp3a33w2u4jvt4

A Dichotomy for Real Boolean Holant Problems [article]

Shuai Shao, Jin-Yi Cai
2020 arXiv   pre-print
We prove a complexity dichotomy for Holant problems on the boolean domain with arbitrary sets of real-valued constraint functions. These constraint functions need not be symmetric nor do we assume any auxiliary functions as in previous results. It is proved that for every set F of real-valued constraint functions, Holant(F) is either P-time computable or #P-hard. The classification has an explicit criterion. This is the culmination of much research on this problem, and it uses previous results
more » ... nd techniques from many researchers. Some particularly intriguing concrete functions f_6, f_8 and their associated families with extraordinary closure properties related to Bell states in quantum information theory play an important role in this proof.
arXiv:2005.07906v1 fatcat:nib2ti4iazgclbfd7umsxodgk4

Laser processing of micro/nano biomimetic structures

Guijian Xiao, Yi He, Shuai Liu, Hao Yi, Lingyan Du
2021 Micro & Nano Letters  
Inspired by nature, many researchers have noticed that the micro-and nano-structures originated from organism's surfaces which indict unique properties such as superhydrophobicity, drag reduction, structure colour have much use in academic value. However, conventional processing methods for mimicking biomimetic surface need multistep and is time wasted, not environmental friendly and uneconomic. Thanks to the development of ultrafast laser technology, the pulse duration can reach femtosecond
more » ... le, which can precisely realize micro-and nano-processing in one-step. Here, we introduce and demonstrate the principles of some typical properties. The typical structures caused by interaction between laser and material are discussed. Moreover, laser-processing methods for biomimetic surface are highlighted to achieve specialized performance.
doi:10.1049/mna2.12055 fatcat:fqezdzvamjeynisvvxf2vvlogu

Morphology Optimization Design Based on Hyperworks Block

Shuai-shuai LIU, Jing LI, Xin WANG, Xin-yi SONG
2017 DEStech Transactions on Materials Science and Engineering  
SHyperWorks software is now widely used in various fields of design and analysis software, it is different from ANSYS and other analysis software, it is more professional and fine in Pre processing of model, through meticulous pre processing technology, we can fully ensure the accuracy in the late optimization design. The optimization design of this thesis is to use the optimization module of Hyperworks software to optimize the design. The Optimization module has topology optimization, size
more » ... mization, shape optimization of the three optimization design module, one is selected as the design analysis, design parameters according to the physical characteristics of the workpiece, including material properties assignment, elastic modulus, density and Poisson's ratio of the input. The parameters setting is complete, analysis of constraints on the artifact and stress condition to determine whether it is displacement constraint or stress constraint, constraint is established again in the workpiece surface or a point exerts a force. Then according to the objective function established the response, the specific constraints limit the value of input to the responses of the dialog bar, and finally establish a working loadsteps and solution, and to obtain the optimum effect diagram, can browse and adjust by Hyperview.
doi:10.12783/dtmse/msce2016/10495 fatcat:mu4y7u6nerc45l5vunvfqsrkf4

From Holant to Quantum Entanglement and Back [article]

Jin-Yi Cai, Zhiguo Fu, Shuai Shao
2020 arXiv   pre-print
Holant problems are intimately connected with quantum theory as tensor networks. We first use techniques from Holant theory to derive new and improved results for quantum entanglement theory. We discover two particular entangled states |Ψ_6〉 of 6 qubits and |Ψ_8〉 of 8 qubits respectively, that have extraordinary and unique closure properties in terms of the Bell property. Then we use entanglement properties of constraint functions to derive a new complexity dichotomy for all real-valued Holant
more » ... roblems containing an odd-arity signature. The signatures need not be symmetric, and no auxiliary signatures are assumed.
arXiv:2004.05706v1 fatcat:b4mzukj765c4tpvyhcxc3ygq7a

Hierarchical Deep Recurrent Architecture for Video Understanding [article]

Luming Tang, Boyang Deng, Haiyu Zhao, Shuai Yi
2017 arXiv   pre-print
This paper introduces the system we developed for the Youtube-8M Video Understanding Challenge, in which a large-scale benchmark dataset was used for multi-label video classification. The proposed framework contains hierarchical deep architecture, including the frame-level sequence modeling part and the video-level classification part. In the frame-level sequence modelling part, we explore a set of methods including Pooling-LSTM (PLSTM), Hierarchical-LSTM (HLSTM), Random-LSTM (RLSTM) in order
more » ... address the problem of large amount of frames in a video. We also introduce two attention pooling methods, single attention pooling (ATT) and multiply attention pooling (Multi-ATT) so that we can pay more attention to the informative frames in a video and ignore the useless frames. In the video-level classification part, two methods are proposed to increase the classification performance, i.e. Hierarchical-Mixture-of-Experts (HMoE) and Classifier Chains (CC). Our final submission is an ensemble consisting of 18 sub-models. In terms of the official evaluation metric Global Average Precision (GAP) at 20, our best submission achieves 0.84346 on the public 50% of test dataset and 0.84333 on the private 50% of test data.
arXiv:1707.03296v1 fatcat:6o5tcwmmvfh67o55cwe6ijlgaq

Magnitude Bounded Matrix Factorisation for Recommender Systems [article]

Shuai Jiang, Kan Li, Richard Yi Da Xu
2018 arXiv   pre-print
Low rank matrix factorisation is often used in recommender systems as a way of extracting latent features. When dealing with large and sparse datasets, traditional recommendation algorithms face the problem of acquiring large, unrestrained, fluctuating values over predictions especially for users/items with very few corresponding observations. Although the problem has been somewhat solved by imposing bounding constraints over its objectives, and/or over all entries to be within a fixed range,
more » ... terms of gaining better recommendations, these approaches have two major shortcomings that we aim to mitigate in this work: one is they can only deal with one pair of fixed bounds for all entries, and the other one is they are very time-consuming when applied on large scale recommender systems. In this paper, we propose a novel algorithm named Magnitude Bounded Matrix Factorisation (MBMF), which allows different bounds for individual users/items and performs very fast on large scale datasets. The key idea of our algorithm is to construct a model by constraining the magnitudes of each individual user/item feature vector. We achieve this by converting from the Cartesian to Spherical coordinate system with radii set as the corresponding magnitudes, which allows the above constrained optimisation problem to become an unconstrained one. The Stochastic Gradient Descent (SGD) method is then applied to solve the unconstrained task efficiently. Experiments on synthetic and real datasets demonstrate that in most cases the proposed MBMF is superior over all existing algorithms in terms of accuracy and time complexity.
arXiv:1807.05515v1 fatcat:htjg76kawne57e35ra5dmdgslu

High expression of XPA confers poor prognosis for nasopharyngeal carcinoma patients treated with platinum-based chemoradiotherapy

Xiang Fu, Jiali Hu, Hong-yu Han, Yi-jun Hua, Ling Zhou, Wen-di Shuai, Wu-ying Du, Chun-mei Kuang, Shuai Chen, Wenlin Huang, Ran-yi Liu
2015 OncoTarget  
In this study, we tried to explore if xeroderma pigmentosum complementation group-A (XPA) expression is likely a prognostic prediction factor for locally advanced nasopharyngeal carcinoma (NPC) patients treated with platinumbased chemoradiotherapy, which was considered to bring chemotherapy-related severe toxicity compared with radiotherapy alone. Firstly, MTT assay revealed that downregulating XPA expression in NPC HONE1 and CNE1 cells decreased IC 50 of cisplatin and sensitized cells to
more » ... tin. XPA expression was detected by immunohistochemistry in cancer tissues from locally advanced NPC patients treated with platinum-based chemoradiotherapy. The relationships between XPA expression and clinicopathologic features, overall survival and progression-free survival of patients were evaluated. The results showed that XPA expression was not associated with clinicopathologic parameters, but was likely an independent prognostic factor for patient survival. High XPA level predicts a poor prognosis, and the prediction values were higher in subgroups of younger, higher EBV antibody titer, or treated with concurrent chemoradiotherapy. Combining XPA levels and T/N classifications, we successfully classified these patients into low, medium and high risk groups for platinum-based chemoradiotherapy. These findings suggest that XPA levels may be a potential predictor of prognosis in locally advanced NPC patients treated with platinum-based chemoradiotherapy, and helpful for selecting patients likely to need and benefit from this treatment in future.
doi:10.18632/oncotarget.4424 pmid:26156020 pmcid:PMC4695073 fatcat:usob2vy5afdvveat6bis26rtpa

Iodobenzene Dicarboxylates as Transferrable Oxygen Sources: Synthesis of α-Oxygenated Ketones from Terminal Aryl Alkynes

Bao-Yi Ren, Bao-Yi Ren, Daokun Zhong, Nan Guo, Weikun Duan, Shuai Song, Xue Yang, Daokun Zhong, Nan Guo, Weikun Duan, Shuai Song, Xue Yang
2016 General Chemistry  
A facile and efficient method for preparation of α-oxygenated ketones has been developed via the reaction of terminal aryl alkynes with various iodobenzene dicarboxylates. When bis(trifluoroacetoxy)iodobenzene (PIFA) was used, α-hydroxy ketones were obtained in high to excellent yields. The hypervalent iodine(III) reagents can function as an electrophile, transferrable O-nucleophile, and an excellent leaving group. The protocol represents a direct, atom-efficient and metal-free conversion of
more » ... ynes into α-oxygenated ketones under mild conditions.
doi:10.21127/yaoyigc20150018 fatcat:ncewagzkcvfhxe3eidih5kodom

Erratum to: Elastic dynamic analysis of synchronous belt drive system using absolute nodal coordinate formulation

Shuai Shuai Jia, Yi Min Song
2015 Nonlinear dynamics  
doi:10.1007/s11071-015-2150-x fatcat:o6kc52nhdvbbthgqutjzmc7qya

Variational Relational Point Completion Network [article]

Liang Pan, Xinyi Chen, Zhongang Cai, Junzhe Zhang, Haiyu Zhao, Shuai Yi, Ziwei Liu
2021 arXiv   pre-print
Real-scanned point clouds are often incomplete due to viewpoint, occlusion, and noise. Existing point cloud completion methods tend to generate global shape skeletons and hence lack fine local details. Furthermore, they mostly learn a deterministic partial-to-complete mapping, but overlook structural relations in man-made objects. To tackle these challenges, this paper proposes a variational framework, Variational Relational point Completion network (VRCNet) with two appealing properties: 1)
more » ... babilistic Modeling. In particular, we propose a dual-path architecture to enable principled probabilistic modeling across partial and complete clouds. One path consumes complete point clouds for reconstruction by learning a point VAE. The other path generates complete shapes for partial point clouds, whose embedded distribution is guided by distribution obtained from the reconstruction path during training. 2) Relational Enhancement. Specifically, we carefully design point self-attention kernel and point selective kernel module to exploit relational point features, which refines local shape details conditioned on the coarse completion. In addition, we contribute a multi-view partial point cloud dataset (MVP dataset) containing over 100,000 high-quality scans, which renders partial 3D shapes from 26 uniformly distributed camera poses for each 3D CAD model. Extensive experiments demonstrate that VRCNet outperforms state-of-theart methods on all standard point cloud completion benchmarks. Notably, VRCNet shows great generalizability and robustness on real-world point cloud scans.
arXiv:2104.10154v1 fatcat:ngpstcm7pnf2nlixfjkm4xegh4

Sentence Compression via DC Programming Approach [article]

Yi-Shuai Niu, Xi-Wei Hu, Yu You, Faouzi Mohamed Benammour, Hu Zhang
2019 arXiv   pre-print
Sentence compression is an important problem in natural language processing. In this paper, we firstly establish a new sentence compression model based on the probability model and the parse tree model. Our sentence compression model is equivalent to an integer linear program (ILP) which can both guarantee the syntax correctness of the compression and save the main meaning. We propose using a DC (Difference of convex) programming approach (DCA) for finding local optimal solution of our model.
more » ... mbing DCA with a parallel-branch-and-bound framework, we can find global optimal solution. Numerical results demonstrate the good quality of our sentence compression model and the excellent performance of our proposed solution algorithm.
arXiv:1902.07248v1 fatcat:5dh5ahlssnhwjjngvdgdfpko7m

Forgetting and small G protein Rac

Yichun Shuai, Yi Zhong
2010 Protein & Cell  
(A) Protein & Cell © Higher Education Press and Springer-Verlag Berlin Heidelberg 2010 Yichun Shuai and Yi Zhong Protein & Cell © Higher Education Press and Springer-Verlag Berlin Heidelberg  ...  Conversely, abolishment of the Rac's binding site with PAK efficiently blocks its function in forgetting (Shuai et al., 2010) .  ... 
doi:10.1007/s13238-010-0077-z pmid:21204003 pmcid:PMC4875324 fatcat:wrt3kbxo4vcavn4ospux7tiwyu
« Previous Showing results 1 — 15 out of 3,465 results