Abstract: Representing features at multiple scales is of great importance for numerous vision tasks. Recent advances in backbone convolutional neural networks (CNNs) continually demonstrate stronger ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results