Suggestion - Integrate MobileSAM into the pipeline for lightweight and faster inference

[mdimtiazh]

Reference: https://github.com/ChaoningZhang/MobileSAM

Our project performs on par with the original SAM and keeps exactly the same pipeline as the original SAM except for a change on the image encode, therefore, it is easy to Integrate into any project.

MobileSAM is around 60 times smaller and around 50 times faster than original SAM, and it is around 7 times smaller and around 5 times faster than the concurrent FastSAM. The comparison of the whole pipeline is summarzed as follows:

Best Wishes,

Qiao

[cclamd]

this my project combile with MobileSAM https://github.com/cclamd/MobileSAM-Adapter

[WuJunde]

Hi @mdimtiazh , we are going to combine MobileSAM and MobileSAM-V2 into our repo. Could you give us your contact so we may can have a short discussion or something?

@cclamd, awesome job on your contribution👍! Could you consider bringing it to Med-SAM-Adapter? Med-SAM-Adapter is now under Kids w/ Tokens, a public, non-profit, and open-source community. We're documenting everything transparently, so your input will shine for all to see (check our news update). Your contribution will definitely impact more people if we work together!🤗 Even if you just pop in to share some dev experience, we'd be stoked. Anyway, respect to your effort🫡.

[WuJunde]

done, check our news

[cclamd]

@WuJunde hi! I have been very busy last few month and didn't check email , It's my pleasure to bringing my small project to Med-SAM-Adapter and work together !

Best Wishes!
Cai