1 code implementation • 28 Nov 2022 • Zhengqi Liu, Jie Gui, Hao Luo
Most previous works mask patches of the image randomly, which underutilizes the semantic information that is beneficial to visual representation learning.
no code implementations • 21 Sep 2021 • Yuwen Li, Zhengqi Liu, Ludmil Zikatanov
The training process of neural networks usually optimize weights and bias parameters of linear transformations, while nonlinear activation functions are pre-specified and fixed.