Raif Olson
Raif Olson
My guess is that you would want to make a model that is similar to the Fast ReID model (but more powerful and accurate) to get the embedding vectors for...
I'm guessing that it just doesn't learn very much. Working on a project using JEPA to train some models, so will try it and let you know.
Hi @SkalskiP and @XiaJunfly I know what the issue is. In this PR #1035 I changed the tracker to return the original detections only if they were tracked. It worked...
I'm pretty sure the reason it shows fp32 numbers in the parameters is because when training the network, you need to use the original floating point values for backprop (1.58...
Yea it is also related to #1320, sorry about that. On Wed, Jul 17, 2024 at 6:19 AM Piotr Skalski ***@***.***> wrote: > Hi @rolson24 👋🏻, we have two issues...
Ok I believe I have fixed the issue. The problem was that each track was being re-assigned a new id when it was found after being lost. There also was...
Just wanted to bump this. I think I fixed the issues we were running into.
Good to hear!
@abelBEDOYA, This is interesting, what version of supervision are you using? I seem to remember this was an issue we fixed a few months ago, but it may not be...
Hmm, the latest release is 0.22.0, please try the latest one and see if it helps. In the meantime I will test your code.