How did you get the visual prompts?

kevin-ssy / CLIP_as_RNN

Official Implementation for CVPR 2024 paper: CLIP as RNN: Segment Countless Visual Concepts without Training Endeavor

89 stars 3 forks source link

Closed EricLina closed 7 months ago

EricLina commented 7 months ago

Thank you for your excellent work!

How did you get the visual prompts? Are prompts learnable or just annotated by human?

I am appreciate it if you could solve my questions.

kevin-ssy commented 7 months ago

Hi,

Thanks for your interest in our paper. We have provided the code for generating visual prompts in the paper. Please refer to Algorithm B on page 19.