Zobrazeno 1 - 10
of 17
pro vyhledávání: '"Songlin Dong"'
Autor:
Songlin Dong, Yihong Gong, Jingang Shi, Miao Shang, Xiaoyu Tao, Xing Wei, Xiaopeng Hong, Tiangang Zhou
Publikováno v:
IEEE Transactions on Neural Networks and Learning Systems. :1-15
Publikováno v:
IEEE Transactions on Neural Networks and Learning Systems. :1-14
In this article, we focus on a new and challenging decentralized machine learning paradigm in which there are continuous inflows of data to be addressed and the data are stored in multiple repositories. We initiate the study of data-decentralized cla
Publikováno v:
Proceedings of the AAAI Conference on Artificial Intelligence. 35:1255-1263
In this paper, we focus on the challenging few-shot class incremental learning (FSCIL) problem, which requires to transfer knowledge from old tasks to new ones and solves catastrophic forgetting. We propose the exemplar relation distillation incremen
Autor:
Jingang Shi, Yusi Wang, Songlin Dong, Xiaopeng Hong, Zitong Yu, Fei Wang, Changxin Wang, Yihong Gong
Publikováno v:
Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence.
Face Super-resolution (FSR) task works for generating high-resolution (HR) face images from the corresponding low-resolution (LR) inputs, which has received a lot of attentions because of the wide application prospects. However, due to the diversity
Publikováno v:
IEEE transactions on neural networks and learning systems.
Deep models have shown to be vulnerable to catastrophic forgetting, a phenomenon that the recognition performance on old data degrades when a pre-trained model is fine-tuned on new data. Knowledge distillation (KD) is a popular incremental approach t
Publikováno v:
International Journal of Hydrogen Energy. 45:7444-7454
The incompatibility between nanofillers and polymer, caused by the agglomeration of nanoparticles and their weak interaction with each other, is still a challenge to develop mixed matrix composite membrane. Herein, we introduced the ZIF-8-TA nanopart
Publikováno v:
IEEE Transactions on Circuits and Systems for Video Technology. :1-1
Publikováno v:
ACM Multimedia Asia.
Publikováno v:
Journal of Membrane Science. 581:195-213
Membrane separation process is one of the most promising methods for post-combustion CO2 capture from coal-fired power plant. In this paper, a multicomponent gas separation process using spiral-wound membrane modules is explored to optimize its membr
Publikováno v:
CVPR
The ability to incrementally learn new classes is crucial to the development of real-world artificial intelligence systems. In this paper, we focus on a challenging but practical few-shot class-incremental learning (FSCIL) problem. FSCIL requires CNN
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::988b28bcb99fffcf04368a70cf3f7bd1
http://arxiv.org/abs/2004.10956
http://arxiv.org/abs/2004.10956