Onnx_Runtime_Mobile_ODSC_Tutorial

Overview

Tutorial for ODSC East 2021, "ML Inference on Mobile Device With Onnx Runtime"

• Slides for this tutorial

This demo Android image classification app is loosely based on Google CodeLabs - Getting Started with CameraX, with all the UI components, image analysis class, pre/post processing functionalities ready to use.

In this workshop, we will integrate Onnx Runtime Mobile and MobileNet V2 into this demo app to perform image classification.

Prerequisites

• Android Studio 4.1+

• A modern Android device with a camera

• The prebuilt onnxruntime arm64 AAR package, model and label files available here

• [optional] If you prefer to build your custom Onnx Runtime Mobile Package

  • Onnx Runtime and its prerequisites

Getting Started

1. Clone this repository
2. Open using Android Studio
3. Click Build -> Make Project
4. Run the App on your Android device
5. Let's start adding image classification capability to this app using Onnx Runtime Mobile