Revisiting the Integration of Convolution and Attention for Vision Backbone

Open in new window