LinasKo
LinasKo
Can confirm that I can access depth using my prior code snippet. Note, the following is a note to the devs - I **don't** recommend following blindly if you're looking...
Pose detection comes with an extra complication - hierarchy. Where more than 1 object is detected, you need to preserve the relations between their own keypoints. As little as it...
Tiny tidbit for the future, as I've done a project with Mediapipe before: using `visibility` is a great starting point for `confidence` as it isn't provided, but there's cases (e.g....
I found issues in data loaders of `sv.Keypoints` when no detections are present or objects are detected (this one in `from_ultralytics` only). Feel free to include it, if you find...
Sure!
We're reviving this one. Some context for new contributors: > SkalskiP: > The InferenceSlicer consists of two key elements: a moving window that slices the image into smaller, partially overlapping...
@sharingan000, how does this sound? Would you have the time to help out? I'm here to help if needed :wink:
Me and **SkalskiP** had a conversation about this - I'll take over for now.
Intermediate results: 1. I've confirmed that threads help, especially when the model is run on the CPU. I see a 5-10x performance improvement. 2. I've implemented the batched inference slicer,...
https://colab.research.google.com/drive/1j85QErM74VCSLADoGliM296q4GFUdnGM?usp=sharing As you can see, in these tests it only helped the Ultralytics case. Known insufficiencies: * ~~Inference 1 model is fit for vehicle detection but is tested on an...