Pose Estimation - Ultralytics YOLO11

This sample shows how to run the Ultralytics YOLO11 pose estimation model on BPU based on hbm_runtime, enabling human keypoint detection and visualization. It supports model preprocessing, inference execution, and postprocessing (including keypoint decoding, bounding box drawing, and keypoint annotation). The sample code is located in /app/pydev_demo/04_pose_sample/01_ultralytics_yolo11_pose/.

This sample shows how to run the Ultralytics YOLO11 pose estimation model on BPU based on hbm_runtime, enabling human keypoint detection and visualization. It supports model preprocessing, inference execution, and postprocessing (including keypoint decoding, bounding box drawing, and keypoint annotation). The sample code is located in /app/pydev_demo/pose_sample/ultralytics_yolo11_pose/.

Model Description

• Introduction:

  Ultralytics YOLO11 Pose is an efficient lightweight human keypoint detection model that supports simultaneous object detection and pose estimation (multi-keypoint prediction). It integrates Distribution Focal Loss (DFL) to improve bounding box and keypoint localization accuracy, making it suitable for multi-person pose recognition tasks in real-time applications.

• HBM model name: yolo11n_pose_nashe_640x640_nv12.hbm

• HBM model name: yolo11n_pose_nashp_640x640_nv12.hbm

• Input format: NV12 image (Y and UV separated), size 640×640

• Output:

  • Bounding box coordinates for each person (xyxy)

  • Keypoint positions (K×2, x/y coordinates)

  • Confidence score for each keypoint

  • Supports COCO human keypoint format (commonly 17 keypoints)

• Model download URL (automatically downloaded by the program):

  https://archive.d-robotics.cc/downloads/rdk_model_zoo/rdk_s100/ultralytics_YOLO/yolo11n_pose_nashe_640x640_nv12.hbm

  https://archive.d-robotics.cc/downloads/rdk_model_zoo/rdk_s600/ultralytics_YOLO/yolo11n_pose_nashp_640x640_nv12.hbm

Features

• Model loading

  Use hbm_runtime to load the specified Ultralytics YOLO11 pose estimation model and automatically parse input/output tensor names, shapes, and quantization parameters.

• Input preprocessing

  Resize the input BGR image to 640×640 and convert it to NV12 format (Y and UV separated) for model inference.

• Inference execution

  Execute inference through the .run() interface. Running priority and BPU core binding can be configured via set_scheduling_params().

• Output postprocessing

  • Decode bounding boxes from multi-scale outputs (using DFL bin decoding);

  • Decode keypoint positions and confidence (K×2 + K);

  • Apply NMS to remove redundant detection boxes;

  • Map keypoint coordinates and bounding boxes back to the original image size;

  • Threshold control to display only high-confidence keypoints;

  • Image visualization support, including detection boxes and keypoint drawing.

Environment Dependencies

This sample has no special environment requirements. You only need to ensure the dependencies in pydev are installed.

pip install -r ../../requirements.txt

pip install -r ../../requirements.txt --break-system-packages

Directory Structure

.
├── ultralytics_yolo11_pose.py    # Main inference script
└── README.md                     # Usage instructions

Parameter Description

Parameter	Description	Default Value
--model-path	Model file path (.hbm format)	/opt/hobot/model/s100/basic/yolo11n_pose_nashe_640x640_nv12.hbm
--test-img	Test image path	/app/res/assets/bus.jpg
--label-file	Class label path, one class name per line	/app/res/labels/coco_classes.names
--img-save-path	Detection result save path	result.jpg
--priority	Model scheduling priority (0~255; larger is higher)	0
--bpu-cores	List of BPU core IDs for inference (for example, --bpu-cores 0 1)	[0]
--nms-thres	IoU threshold for non-maximum suppression (NMS)	0.7
--score-thres	Object confidence threshold (objects below this value are filtered)	0.25
--kpt-conf-thres	Keypoint visualization confidence threshold (keypoints below this value are not displayed)	0.5

Parameter	Description	Default Value
--model-path	Model file path (.hbm format)	/opt/hobot/model/s600/basic/yolo11n_pose_nashp_640x640_nv12.hbm
--test-img	Test image path	/app/res/assets/bus.jpg
--label-file	Class label path, one class name per line	/app/res/labels/coco_classes.names
--img-save-path	Detection result save path	result.jpg
--priority	Model scheduling priority (0~255; larger is higher)	0
--bpu-cores	List of BPU core IDs for inference (for example, --bpu-cores 0 1)	[0]
--nms-thres	IoU threshold for non-maximum suppression (NMS)	0.7
--score-thres	Object confidence threshold (objects below this value are filtered)	0.25
--kpt-conf-thres	Keypoint visualization confidence threshold (keypoints below this value are not displayed)	0.5

Quick Start

• Run the model

  • Use default parameters

    python ultralytics_yolo11_pose.py

  • Run with specified parameters

    python ultralytics_yolo11_pose.py \
    --model-path /opt/hobot/model/s100/basic/yolo11n_pose_nashe_640x640_nv12.hbm \
    --test-img /app/res/assets/bus.jpg \
    --label-file /app/res/labels/coco_classes.names \
    --img-save-path result.jpg \
    --priority 0 \
    --bpu-cores 0 \
    --score-thres 0.25 \
    --nms-thres 0.7 \
    --kpt-conf-thres 0.5

    python ultralytics_yolo11_pose.py \
    --model-path /opt/hobot/model/s600/basic/yolo11n_pose_nashp_640x640_nv12.hbm \
    --test-img /app/res/assets/bus.jpg \
    --label-file /app/res/labels/coco_classes.names \
    --img-save-path result.jpg \
    --priority 0 \
    --bpu-cores 0 \
    --score-thres 0.25 \
    --nms-thres 0.7 \
    --kpt-conf-thres 0.5

• View the result

  After successful execution, results will be drawn on the original image and saved to the path specified by --img-save-path.

  [Saved] Result saved to: result.jpg

Notes

• If the specified model path does not exist, the program will attempt to download the model automatically.

License

Copyright (C) 2025, XiangshunZhao D-Robotics.

This program is free software: you can redistribute it and/or modify
it under the terms of the GNU Affero General Public License as
published by the Free Software Foundation, either version 3 of the
License, or (at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU Affero General Public License for more details.

You should have received a copy of the GNU Affero General Public License
along with this program.  If not, see <https://www.gnu.org/licenses/>.