21–30 of 32 posts
I have no idea what this even is.
Earlier quoted context omitted.
> An extended family of versatile vision foundation models producing high-quality dense features and achieving outstanding performance on various vision tasks including outperforming the specialized state of the art across a broad range of settings, without fine-tuning
To elaborate, this is a foundation model. This basically means it can take an arbitrary image and map it to a high dimensional space H in which ~arbitrary characteristics become much easier to solve for. For example (and this might be oversimplifying a bit, computer vision people please correct me if I’m wrong) if you’re interested in knowing whether or not the image contains a cat, then maybe there is some hyperplan…
> ViT models pretrained on satellite dataset (SAT-493M)
DINOv2 had pretty poor out-of-the-box performance on satellite/aerial imagery, so it's super exciting that they released a version of it specifically for this use case.
I’m fascinated by this, but am admittedly clueless about how to actually go about building any kind of recognizer or other system atop it.
Could anyone point to an example or git repo showing a simple implementation? I’m fascinated by this, but am admittedly clueless about how to actually go about building any kind of recognizer or other system atop it.
As for doing it in general, it's a fairly standard vision transformer so anything built on DINOv2 (or any other ViT) should be easy to adapt to v3.
Earlier quoted context omitted.
This has nothing to do with the newly appointed fellow nor Meta Superintelligence Labs, but rather work from FAIR that would have gone through a lengthy review process before seeing the light of day. Not fun to see the license change in any case
I remember DINOv2 was originally a commercial licence. I (along with others) just asked if they could change it on a GitHub issue, and after some time, they did. Might be worth asking
Earlier quoted context omitted.
> An extended family of versatile vision foundation models producing high-quality dense features and achieving outstanding performance on various vision tasks including outperforming the specialized state of the art across a broad range of settings, without fine-tuning
To elaborate, this is a foundation model. This basically means it can take an arbitrary image and map it to a high dimensional space H in which ~arbitrary characteristics become much easier to solve for. For example (and this might be oversimplifying a bit, computer vision people please correct me if I’m wrong) if you’re interested in knowing whether or not the image contains a cat, then maybe there is some hyperplan…
Could anyone point to an example or git repo showing a simple implementation? I’m fascinated by this, but am admittedly clueless about how to actually go about building any kind of recognizer or other system atop it.
[0]: https://github.com/tue-mps/eomt [1]: https://docs.lightly.ai/train/stable/semantic_segmentation.h...
Earlier quoted context omitted.
To elaborate, this is a foundation model. This basically means it can take an arbitrary image and map it to a high dimensional space H in which ~arbitrary characteristics become much easier to solve for. For example (and this might be oversimplifying a bit, computer vision people please correct me if I’m wrong) if you’re interested in knowing whether or not the image contains a cat, then maybe there is some hyperplan…
Thanks, I think I understand roughly. Could it be used for for recognizing people? As in identifying what person is in what image?
I imagine it would depend on whether DINOv3 captures the information of whether a given person is in the image, which I think is really a question about training data. So naively, I would guess the answer is yes for celebrities and no for non-celebrities. Partially for data/technical reasons, but also maybe due to the murkier legal expectation of privacy for famous people.