You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hey,
Thank you for your fantastic work! I have a question about Seg-GPT. Is it possible to use multiple prompts, meaning pairs of images and masks, instead of just one pair? Initially, I thought about concatenating several images and their corresponding masks and feeding them to the model as a single unified prompt. However, due to resolution limitations, this approach tends to degrade the quality of the prompt. Currently, I'm unsure of how to provide the model with a perfect and relatively general prompt.
The text was updated successfully, but these errors were encountered:
Hey,
Thank you for your fantastic work! I have a question about Seg-GPT. Is it possible to use multiple prompts, meaning pairs of images and masks, instead of just one pair? Initially, I thought about concatenating several images and their corresponding masks and feeding them to the model as a single unified prompt. However, due to resolution limitations, this approach tends to degrade the quality of the prompt. Currently, I'm unsure of how to provide the model with a perfect and relatively general prompt.
The text was updated successfully, but these errors were encountered: