IDEA-Research / Grounded-SAM-2

Grounded SAM 2: Ground and Track Anything in Videos with Grounding DINO, Florence-2 and SAM 2
https://arxiv.org/abs/2401.14159
Apache License 2.0
708 stars 49 forks source link

Making SAM 2 run 2x faster #37

Open mvoodarla opened 2 weeks ago

mvoodarla commented 2 weeks ago

I was pretty amazed with SAM 2 when it came out given all the work I do with video. My company works a ton with it and we decided to take a crack at optimizing it, and we made it run 2x faster than the original pipeline!

Unlike LLMs, video models are notorious for incredibly inefficient file reading, storage, and writing which makes them much slower than they need to be.

We wrote a bit about our work here and thought we'd share with the community: https://www.sievedata.com/blog/meta-segment-anything-2-sam2-introduction

rentainhe commented 2 weeks ago

Awesome work! It will be much more convenient for community users to annotate their own datasets.

MonolithFoundation commented 1 week ago

@mvoodarla Are there merely advertisements for your cloud services? No code? Additionally, I didn't see any substantive content in the blog.

mvoodarla commented 1 week ago

Hey! We're doing a follow up blog soon with technical details as well.