Zobrazeno 1 - 10
of 4 370
pro vyhledávání: '"Gu, Li"'
Autor:
Wang, Ziqiang, Chi, Zhixiang, Wu, Yanan, Gu, Li, Liu, Zhi, Plataniotis, Konstantinos, Wang, Yang
Given a model trained on source data, Test-Time Adaptation (TTA) enables adaptation and inference in test data streams with domain shifts from the source. Current methods predominantly optimize the model for each incoming test data batch using self-t
Externí odkaz:
http://arxiv.org/abs/2407.12128
Autor:
Chi, Zhixiang, Gu, Li, Zhong, Tao, Liu, Huan, Yu, Yuanhao, Plataniotis, Konstantinos N, Wang, Yang
In this paper, we aim to adapt a model at test-time using a few unlabeled data to address distribution shifts. To tackle the challenges of extracting domain knowledge from a limited amount of data, it is crucial to utilize correlated information from
Externí odkaz:
http://arxiv.org/abs/2405.02797
Autor:
Xiao-Hui Deng, Long-Xin Huang, Qi- Sun, Chan-Gu Li, Ying-Chun Xie, Xiao-Qing Liu, Qing-Ling Fu
Publikováno v:
Heliyon, Vol 10, Iss 17, Pp e36218- (2024)
Background: Low-density neutrophils are heterogeneous immune cells with immunosuppressive (such as polymorphonuclear myeloid-derived suppressor cells [PMN-MDSC]) or pro-inflammatory (such as low-density granulocytes [LDG]) properties that have been w
Externí odkaz:
https://doaj.org/article/bc779ea3e9dc45508cec3ee16ca004fa
Publikováno v:
Swiss Journal of Paleontology. 9/18/2024, Vol. 143 Issue 1, p1-16. 16p.
Publikováno v:
Journal of Functional Materials / Gongneng Cailiao. 2024, Vol. 55 Issue 9, p09007-09021. 8p.
In this paper, we tackle the problem of domain shift. Most existing methods perform training on multiple source domains using a single model, and the same trained model is used on all unseen target domains. Such solutions are sub-optimal as each targ
Externí odkaz:
http://arxiv.org/abs/2210.03885
In this work, we present the winning solution for ORBIT Few-Shot Video Object Recognition Challenge 2022. Built upon the ProtoNet baseline, the performance of our method is improved with three effective techniques. These techniques include the embedd
Externí odkaz:
http://arxiv.org/abs/2210.00174
Few-shot class-incremental learning (FSCIL) has been proposed aiming to enable a deep learning system to incrementally learn new classes with limited data. Recently, a pioneer claims that the commonly used replay-based method in class-incremental lea
Externí odkaz:
http://arxiv.org/abs/2207.11213
Autor:
Yang, Tian, Liu, Da-Qun, Qiu, Wei, Fan, Zhong-Qi, Sun, Li-Yang, Wang, Nan-Ya, Wang, Hong, Yang, Yi-Fan, Li, Jie, Zhou, Ya-Hao, Chen, Ting-Hao, Wang, Xian-Ming, Gu, Wei-Min, Liang, Ying-Jian, Gu, Li-Hui, Xu, Jia-Hao, Wang, Ming-Da, Sun, Xiao-Dong, Lv, Guo-Yue
Publikováno v:
In The American Journal of Surgery November 2024 237
Publikováno v:
In Science of the Total Environment 20 October 2024 948