search

youweiliang / evit

Python code for ICLR 2022 spotlight paper EViT: Expediting Vision Transformers via Token Reorganizations
Apache License 2.0
162 stars 19 forks source link

The figure #27

Open wyl753 opened 6 months ago

wyl753 commented 6 months ago

Awesome works! I want to know how Figure 5 is generated.Is it by putting the mask back on the original image ?