yformer / EfficientSAM

EfficientSAM: Leveraged Masked Image Pretraining for Efficient Segment Anything
Apache License 2.0
2.1k stars 149 forks source link

How to use a text prompt. #26

Open GitChanyoung opened 9 months ago

GitChanyoung commented 9 months ago

Thank you for your amazing research. I would like to use a text prompt. However, when I examined the code, there was no code related to the text prompt. I'm curious if you have any plans to add it in the future or if there's a specific reason why text prompts are not available. Thank you.

yformer commented 9 months ago

@GitChanyoung, thanks for your interest in EfficientSAM! You can check Grounded EfficientSAM demo on Grounded-Efficient-Segment-Anything, which you can perform text prompt.

Here is one example from Grounded-Efficient-Segment-Anything,

Grounded-EfficientSAM
xiaobanni commented 9 months ago

Grounded-Segment-Anything is a two-stage process, and not the direct and elegant solution from text to Segment as depicted in the original Segment Anything, as discussed in the following link. I believe text prompts are a broad demand, and look forward to subsequent work focusing on text prompts.

yformer commented 9 months ago

@xiaobanni, Yes, Grounded-SAM is a two-stage process. We will consider the text prompt direction. Thanks for your interest.