Dynamic Grained Encoder for Vision Transformers A Limitation and Future Work

The budget for DGE is set to 0.5. "Resolution" refers to the side length of input images. As shown in Figure 1(a), one limitation of our work is that the acceleration ratio on GPUs (based on native PyTorch implementation) is not good when the input image size is small. The metric in [64] is used to measure the gating scores in each DGE layer. As shown in Tab. 5 and Tab.