icml2023

Global context vision transformers
Global context vision transformers
New vision transformer architecture optimized for number of parameters and FLOPs.
A. Hatamizadeh, H. Yin, J. Kautz, P. Molchanov
PDF Cite Code ICML2023