MSViT: Dynamic Mixed-Scale Tokenization for Vision Transformers

Add code
Jul 05, 2023

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: