Abstract: Recent progress on Transformers and multilayer perceptron (MLP) models provide new network architectural designs for computer vision tasks. Although these models proved to be effective in ...