Skip to content

lmoroney/android-beans-demo

Repository files navigation

android-beans-demo

Demo Step 1:

Create an android app. Incorporate ML Kit and CameraX. Use CameraX to grab an image, pass it to MLKit, get the labels back, and render them in a toast.

** This is implemented as beandisease_poc android app **

Demo Step 2:

Show how to use TensorFlow Model Maker to create a custom model using the beans dataset in TensorFlow Data Services.

** This is implemented as Beans_With_ModelMaker.ipynb **

Demo Step 3:

Incorporate the beans model into the app, using Import->Other->TensorFlow Lite. Updae the code to use CameraX to get the image and pass it to this model. Get the labels back and render them in a toast

** This is implemented as beandisease_poc_lite android app **

Demo Step 4

Add Firebase services to the app, showing how, when moving towards productizing an app you may need things like analytics, sign-in or remote config

** This is implemented as beandisease_pod_lite_firebase android app **

Demo Step 5

Show the beans model running in the browser with TensorFlow.js

** This is implemented as beansapp **

About

Stuff for the android ML demo for beans app

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published