search

raoyongming / DenseCLIP

[CVPR 2022] DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting
505 stars 38 forks source link

Prompt learning via CoOp #34

Closed anurag-198 closed 1 year ago

anurag-198 commented 1 year ago

Hi, Thanks for your work. In the paper, it is mentioned that you use CoOp to generate prompts where <p1, p2,...,pn> are learnable. But from the code it seems that only the class labels are used for generating the prompt text ? Can you please confirm this, if I am missing something ?

raoyongming commented 1 year ago

Hi, thanks for your interest in our work. The text prompt used in DenseCLIP consists of the labels prompts defined here and class labels. They are combined and fed into the text encoder to generate the text features here.

anurag-198 commented 1 year ago

Ahh, totally missed the the context prompt learnable parameter :) Thanks for making it simpler and quick response.