search

mit-han-lab / efficientvit

EfficientViT is a new family of vision models for efficient high-resolution vision.
Apache License 2.0
1.6k stars 142 forks source link

Cityscapes dataset train #37

Closed d710055071 closed 8 months ago

d710055071 commented 8 months ago

May I ask how many categories do you use with the Cityscapes dataset, what are they, and is there a corresponding relationship