Hello,
I looked through the model zoo provided in the github repo for voyager-sdk and could not find support for all the yolo11 pose estimation models. Specifically for yolo11m-pose and yolo11x-pose. Is support for these two models something that is being considered in the (near) future? Is it possible to deploy these two models with the repo being as is?
Solved
Yolo11 Pose Estimation Models Support
Best answer by Spanner
Hi
Indeed, I think the model zoo currently includes YOLO11n-pose and YOLO11l-pose. The m and x variants aren't listed, but as far as I know all YOLO11 pose models share the same underlying architecture, so you could be able to deploy them using the custom weights workflow.
The approach would be to take one of the existing YOLO11 pose YAML files (e.g. yolo11l-pose) as your starting point, then swap in the weights for the m or x variant. The custom weights tutorial walks through this. You'd export your chosen model to ONNX (opset 11, with simplify=True and dynamic=False) and update the YAML accordingly.
A community member actually did something very similar with yolov8x (also not in the zoo) by adapting the yolov8l YAML, and it worked.
What’s the project you’re working on?
+3- Axelera Team
- Answer
Hi
Indeed, I think the model zoo currently includes YOLO11n-pose and YOLO11l-pose. The m and x variants aren't listed, but as far as I know all YOLO11 pose models share the same underlying architecture, so you could be able to deploy them using the custom weights workflow.
The approach would be to take one of the existing YOLO11 pose YAML files (e.g. yolo11l-pose) as your starting point, then swap in the weights for the m or x variant. The custom weights tutorial walks through this. You'd export your chosen model to ONNX (opset 11, with simplify=True and dynamic=False) and update the YAML accordingly.
A community member actually did something very similar with yolov8x (also not in the zoo) by adapting the yolov8l YAML, and it worked.
What’s the project you’re working on?
Thanks for the reply!
I'll look into trying something similar to what you suggested. To answer your question, not really working on something specific just doing some research and trying to figure out how some of this stuff works.
Sign up
Already have an account? Login
Log in, or create an Axelera AI account
Log In or Register HereEnter your E-mail address. We'll send you an e-mail with instructions to reset your password.