ARCore for Android & Tensorflow Lite

Getting Started - ARCore for Android & Tensorflow Lite

Launching for AR!! This is going to be a series of AR learning notes, it’s about sharing and learning.

ARCore for Android

Generally, I followed the official guidelines to setup the environment and run the samples.

ARCore - Quickstart for Android

Note that if you don’t have an Android device, please turn to the second part in the link above first and install a higher version of Android studio.



For me, I need to run ARCore in an emulator so I installed Android Studio 3.1 RC 2. (I didn’t install Canary 6 because I thought 3.1 RC might be a little more stable than the latest one.)

Just follow the instructions and have fun.

Tips: As you move your device, the app automatically detects flat surfaces and draws a diamond grid. Tap the grid to place an Android robot on the surface.

Tensorflow Lite

It takes a while for me to succesfully run tensorflow

If you fail, there could be tons of reasons in every single step. But it’s nothing hard if you just stick to the guidelines.

1. install tensorflow

Follow the instructions here

For ubuntu users [install tensorflow on ubuntu]

For windows users [install tensorflow on windows]

For macOS users [install tensorflow on Mac]

2. Run android demo app

I recommend that you follow the documentation on GitHub instead of the one on website to run demo android app, as it’s clearer about certain details, at least for me in this case.

On GitHub github documentation-tensorflow

On website tensorflow-getting started

注： 如果想看中文的朋友可以留言我以后会加中文！