HoloLens2-Machine-Learning - GitLite

KO | EN

gitlite — search

#javascript #python #react #typescript #deep-learning #linux #docker #macos #nodejs #windows #awesome #vue #pytorch #openai

HoloLens2-Machine-Learning

(★ 113)

Using deep learning models for image classification directly on the HoloLens 2.

파일 탐색기

최종 버전 다운로드 (.zip)

- - Labels.json.txt
  - Labels.json.txt.meta
- - inference.unity
  - inference.unity.meta
- - MediaCaptureUtility.cs
  - MediaCaptureUtility.cs.meta
  - model.cs
  - model.cs.meta
  - NetworkBehaviour.cs
  - NetworkBehaviour.cs.meta
  - NetworkModel.cs
  - NetworkModel.cs.meta
- Resources.meta
- Scenes.meta
- Scripts.meta
- WSATestCertificate.pfx
- WSATestCertificate.pfx.meta
- model.onnx
- README.md
- manifest.json
- packages-lock.json
- AudioManager.asset
- ClusterInputManager.asset
- DynamicsManager.asset
- EditorBuildSettings.asset
- EditorSettings.asset
- GraphicsSettings.asset
- InputManager.asset
- NavMeshAreas.asset
- PackageManagerSettings.asset
- Physics2DSettings.asset
- PresetManager.asset
- ProjectSettings.asset
- ProjectVersion.txt
- QualitySettings.asset
- TagManager.asset
- TimeManager.asset
- UnityConnectSettings.asset
- VFXManager.asset
- XRSettings.asset
- french-bulldog.jpg
.gitignore
french-bulldog-detection.jpg
LICENSE
onnx-model-load.PNG
README.md

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# HoloLens-2-Machine-Learning
Using the `EfficientNetB0` model, trained on the `ImageNet` 1000 class dataset, for image classification. Model inference is run directly on the HoloLens 2 using its onboard CPU.

## About
- Optimal performance is achieved using version 19041 builds. In this sample I am using build 19041.1161 (Windows Holographic, version 20H2 - August 2021 Update) which can be downloaded from MSFT via the following [link](https://aka.ms/hololens2download/10.0.19041.1161) and installed using the [Advanced Recovery Companion](https://www.microsoft.com/en-ca/p/advanced-recovery-companion/9p74z35sfrs8?rtc=1&activetab=pivot:overviewtab)
- Tested with Unity 2019.4 LTS, Visual Studio 2019, and the HoloLens 2
- Building off of the [WinMLExperiments](https://github.com/reneschulte/WinMLExperiments) sample from Rene Schulte
- Input video frames of size `(224, 224)` for online inference
- Pretrained TensorFlow-Keras implementation of the EfficientNetB0 framework was converted directly to ONNX format for use in this sample

## Run sample
- Open sample in Unity
- Switch build platform to `Universal Windows Platform`, select `HoloLens` for target device, and `ARM64` as the target platform
- Build Visual Studio project and open .sln file
- Copy the `onnx-models\model.onnx` file to the `Builds\HoloLens-2-Machine-Learning\Assets` folder
- Import to Visual Studio project as an existing file, place in the assets folder
- In the asset properties window (as below), confirm that the `Content` field has its boolean value set to `True`. This enables the `ONNX` model to be loaded at runtime from the Visual Studio assets folder

![](https://raw.githubusercontent.com/doughtmw/HoloLens2-Machine-Learning/main/onnx-model-load.PNG)

- Build the sample in `Release` mode for `ARM64` and deploy to the HoloLens 2 to test
- Prediction labels are pulled from the parsed ImageNet labels .json file (which includes 1000 image net classes)
- Output includes the `predicted label`, its associated `probability` and the `inference time` in milliseconds

![](https://raw.githubusercontent.com/doughtmw/HoloLens2-Machine-Learning/main/french-bulldog-detection.jpg)

## Model conversion to ONNX
- See sample conversion from official TensorFlow efficientnet weights to ONNX format in [README file](onnx-models/README.md)