PoseDetectionBarracuda is a human pose detecter that runs the Mediapipe Pose Detection neural network model on the Unity Barracuda.

PoseDetectionBarracuda implementation is inspired by BlazePalmBarracuda and I referenced his source code.(Thanks, keijiro!).

PoseDetectionBarracuda can be installed with npm or GitHub URL.

PoseDetectionBarracuda can be installed by adding following sections to the manifest file (Packages/manifest.json).

To the scopedRegistries section:

{
  "name": "creativeikep",
  "url": "https://registry.npmjs.com",
  "scopes": [ "jp.ikep" ]
}

To the dependencies section:

"jp.ikep.mediapipe.posedetection": "1.0.1"

Finally, the manifest file looks like below:

{
    "scopedRegistries": [
        {
            "name": "creativeikep",
            "url": "https://registry.npmjs.com",
            "scopes": [ "jp.ikep" ]
        }
    ],
    "dependencies": {
        "jp.ikep.mediapipe.posedetection": "1.0.1",
        ...
    }
}

PoseDetectionBarracuda can be installed by adding below URL on the Unity Package Manager's window.

https://github.com/creativeIKEP/PoseDetectionBarracuda.git?path=Packages/PoseDetectionBarracuda#v1.0.1

This image for demo scene("SampleScene.unity") was downloaded from here.

The ONNX model files have been converted for Unity Barracuda from Mediapipe's "pose_detection.tflite" file. The conversion operation is the same as FaceLandmarkBarracuda by keijiro. Check his operation script for details.

IKEP

Copyright (c) 2021 IKEP

Apache-2.0