Zobrazeno 1 - 10
of 1 954
pro vyhledávání: '"CHEN Yimin"'
Federated learning, while being a promising approach for collaborative model training, is susceptible to poisoning attacks due to its decentralized nature. Backdoor attacks, in particular, have shown remarkable stealthiness, as they selectively compr
Externí odkaz:
http://arxiv.org/abs/2407.09658
Autor:
Chai, Xiaoxiang, Chen, Yimin
We prove a new Minkowski type formula for capillary hypersurfaces supported on totally geodesic hyperplanes in hyperbolic space. It leads to a volume-preserving flow starting from a star-shaped initial hypersurface. We prove the long-time existence o
Externí odkaz:
http://arxiv.org/abs/2405.06934
Autor:
Chen, Yimin, Pyo, Juncheol
In this paper, we prove a Heintze-Karcher type inequality for capillary hypersurfaces supported on various hypersurfaces in the hyperbolic space. The equality case only occurs on capillary totally umbilical hypersurfaces. Then we apply this result to
Externí odkaz:
http://arxiv.org/abs/2206.09062
Autor:
Zhang, Jiawei, Wang, Xiang, Bai, Xiao, Wang, Chen, Huang, Lei, Chen, Yimin, Gu, Lin, Zhou, Jun, Harada, Tatsuya, Hancock, Edwin R.
Despite recent stereo matching networks achieving impressive performance given sufficient training data, they suffer from domain shifts and generalize poorly to unseen domains. We argue that maintaining feature consistency between matching pixels is
Externí odkaz:
http://arxiv.org/abs/2203.10887
Publikováno v:
In NeuroImage 15 October 2024 300
Publikováno v:
In Journal of Alloys and Compounds 15 August 2024 995
Autor:
Gu, Jierong, Chen, Yimin, Shen, Xiang, Jia, Guang, Xia, Kelun, Wu, Miaomiao, Xu, Tiefeng, Liu, Zijun
Publikováno v:
In Infrared Physics and Technology August 2024 140
Publikováno v:
In Building and Environment 1 August 2024 261
Autor:
Xie, Shuangquan, Gu, Jierong, Jia, Guang, Liu, Zijun, Gu, Chenjie, Gao, Yixiao, Zheng, Wenfeng, Liu, Ziqiang, Xu, Tiefeng, Shen, Xiang, Chen, Yimin
Publikováno v:
In Infrared Physics and Technology August 2024 140
Weakly-Supervised Semantic Segmentation (WSSS) segments objects without a heavy burden of dense annotation. While as a price, generated pseudo-masks exist obvious noisy pixels, which result in sub-optimal segmentation models trained over these pseudo
Externí odkaz:
http://arxiv.org/abs/2112.07431