https://github.com/creativeikep/poselandmarkbarracuda
PoseLandmarkBarracuda is a human pose landmark detecter that runs the Mediapipe Pose Landmark neural network model on the Unity Barracuda.
https://github.com/creativeikep/poselandmarkbarracuda
barracuda deep-learning human landmark mediapipe pose unity
Last synced: over 1 year ago
JSON representation
PoseLandmarkBarracuda is a human pose landmark detecter that runs the Mediapipe Pose Landmark neural network model on the Unity Barracuda.
- Host: GitHub
- URL: https://github.com/creativeikep/poselandmarkbarracuda
- Owner: creativeIKEP
- License: apache-2.0
- Created: 2021-05-13T09:29:02.000Z (about 5 years ago)
- Default Branch: main
- Last Pushed: 2022-01-09T12:38:39.000Z (over 4 years ago)
- Last Synced: 2025-02-28T05:56:30.414Z (over 1 year ago)
- Topics: barracuda, deep-learning, human, landmark, mediapipe, pose, unity
- Language: C#
- Homepage:
- Size: 57.7 MB
- Stars: 16
- Watchers: 3
- Forks: 4
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE.md
Awesome Lists containing this project
README
# PoseLandmarkBarracuda
PoseLandmarkBarracuda is a human pose landmark detecter that runs the [Mediapipe Pose](https://google.github.io/mediapipe/solutions/pose) Landmark neural network model on the [Unity Barracuda](https://docs.unity3d.com/Packages/com.unity.barracuda@latest).
PoseLandmarkBarracuda implementation is inspired by [HandLandmarkBarracuda](https://github.com/keijiro/HandLandmarkBarracuda) and I referenced [his](https://github.com/keijiro) source code.(Thanks, [keijiro](https://github.com/keijiro)!).
## Install
PoseLandmarkBarracuda can be installed with npm or GitHub URL.
### Install from npm (Recommend)
PoseLandmarkBarracuda 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.poselandmark": "1.1.1"
```
Finally, the manifest file looks like below:
```
{
"scopedRegistries": [
{
"name": "creativeikep",
"url": "https://registry.npmjs.com",
"scopes": [ "jp.ikep" ]
}
],
"dependencies": {
"jp.ikep.mediapipe.poselandmark": "1.1.1",
...
}
}
```
### Install from GitHub URL
PoseLandmarkBarracuda can be installed by adding below URL on the Unity Package Manager's window
```
https://github.com/creativeIKEP/PoseLandmarkBarracuda.git?path=Packages/PoseLandmarkBarracuda#v1.1.1
```
or, adding below sentence to your manifest file(`Packages/manifest.json`) `dependencies` block.
```
"jp.ikep.mediapipe.poselandmark": "https://github.com/creativeIKEP/PoseLandmarkBarracuda.git?path=Packages/PoseLandmarkBarracuda#v1.1.1"
```
## Demo Image
[This image](/Assets/Image/demoImage.jpg) for demo scene(["SampleScene.unity"](/Assets/Scenes/SampleScene.unity)) was downloaded from [here](https://unsplash.com/photos/72zsd_fnxYc).
## ONNX Model
The ONNX model files have been converted for Unity Barracuda from Mediapipe's ["pose_landmark_full.tflite"](https://github.com/google/mediapipe/blob/v0.8.6/mediapipe/modules/pose_landmark/pose_landmark_full.tflite) and ["pose_landmark_lite.tflite"](https://github.com/google/mediapipe/blob/v0.8.6/mediapipe/modules/pose_landmark/pose_landmark_lite.tflite) file.
The ONNX model files were converted with [tflite2tensorflow](https://github.com/PINTO0309/tflite2tensorflow) and [tf2onnx](https://github.com/onnx/tensorflow-onnx).
## Author
[IKEP](https://ikep.jp)
## LICENSE
Copyright (c) 2021 IKEP
[Apache-2.0](/LICENSE.md)