Segment Anything 2, every mention
14 scenes · ← back to Segment Anything 2
tap a year for its mentions
every year anyone Nikhila Ravi 33Joseph Nelson 11Shawn Wang 10Isaac Robinson 1
Verbatim, from the transcripts: the passages where Segment Anything 2 comes up
SAM 3: The Eyes for AI — Nikhila & Pengchuan (Meta Superintelligence), ft. Joseph Nelson (Roboflow)
- ▶ 11:01 unnamed speaker I guess, like, what was the planning like in SAMHII? 3 times in the scene
- ▶ 24:22 unnamed speaker Uh, this is Sam too. 3 times in the scene
- ▶ 45:38 Nikhila Ravi I think video is a whole nother challenging beast and getting to that fully automated data engine is something that we tried to do in SAM too. 2 times in the scene
- ▶ 57:04 Nikhila Ravi Maybe I can comment on SAM specifically, but in SAM-III we did leverage many of the open source contributions people have made on top of SAM-II.
Best of 2024 in Vision [LS Live @ NeurIPS]
- ▶ 1:28 Isaac Robinson And then we're also going to talk about SAM two, which applies the SAM strategy to video, and then how debtors are the improvements in 2024 to debtors that are making them a Pareto improvement to yellow base models.
Segment Anything 2: Memory + Vision = Object Permanence — with Nikhila Ravi and Joseph Nelson
- ▶ 0:06 Shawn Wang I'm delighted to do Segment Anything Two. 4 times in the scene
- ▶ 4:36 Joseph Nelson And with the introduction of Segment Anything, you can pass an image with SAM one, SAM two videos as well, and get perfect, pixel perfect outlines of most everything inside the images. 2 times in the scene
- ▶ 13:02 Shawn Wang I think it's, it's probably a good time to cut to a quick demo of SAM-II for people who are, who are tuning in for SAM-II and who better to demo SAM-II than Nikki? 6 times in the scene
- ▶ 19:05 Nikhila Ravi And so, because we saw that with SAM too, like the demo was a priority first class citizen from day one. 12 times in the scene
- ▶ 23:48 Joseph Nelson So again, I can give some context into one key area that Nikola, you mentioned earlier, which is SAM has made the decision, both SAM-I and SAM-II, to be class agnostic in terms of its predictions, and that you then have the ability to have… 6 times in the scene
- ▶ 32:00 Joseph Nelson So, like, we've, like, even SAM one, SAM two, and the model was released at four p.m. 3 times in the scene
- ▶ 37:26 Nikhila Ravi We had SAM plus a version of SAM-II that only had mask as input.
- ▶ 41:11 Nikhila Ravi Going to SAM-II, we can also apply SAM-II to images because we can, you know, as I said, treat an image as a video with a single frame. 5 times in the scene
- ▶ 48:35 Nikhila Ravi Sam too is another example. 12 times in the scene