Fragments 2019

State of mobile engineering, state of platforms, hardware and user research.

Tickets

Firebase ML Kit : Machine Learning made easy

Submitted by Harshit Dwivedi (@the-dagger) on Thursday, 24 May 2018

Section: Workshop Technical level: Beginner

Abstract

At I/O 2018, Google released the Firebase ML Kit which creates various exciting opportunities for Android Developers aiming to build smart apps without having to worry about the nitty-gritties of Machine Learning.

The Firebase ML Kit APIs offer features like face detection, text recognition, object detection, etc.
Your apps can also label a provided image for special characteristics and identify popular landmarks in a picture.

In this talk, I will outline the usage of all the 5 APIs available in Firebase ML Kit
and I’ll be doing so by using a sample app that utilizes these APIs.

I will be walking you through the working of each api and you will leave the talk having sufficient knowledge of the APIs to go ahead and implement them in your own apps.

Outline

The talk is a tech talk which will outline the usage of all the 5 APIs available in Firebase ML Kit.
I’ll be using a sample app that I’ve created and will be walking the participants through the working of each api.

I’ll also be talking about how they can upload a custom model to firebase and use that instead of using the preloaded models.

After the talk, the attendees will have a good overview of these newly introduced apis and they will have enough knowledge to go ahead and implement them in their apps.

  1. Who should attend this tutorial?
    Android Developers who are excited by the concept of Machine Learning but don’t have a good idea on how to start with it.
    The talk will start with covering some basic APIs and then move on to how you can train a custom model of your own and use it to perform inferencing in your own app.

  2. Why attend?
    You’ll get a good overview of how you can start experiementing with Machine Learning by combining it with Android Development.
    This workshop will also serve as a motivator for you to get started with Machine Learning.

  3. Duration of the tutorial.
    2.5 hours to 3 hours

  4. Prior requirements in terms of knowledge; software installation and equipment that a participant needs to carry for this tutorial.
    Basics of Android Development, basic Kotlin, Android Stduio, Python3, Tensorflow (optional)

Some blogs I’ve written on the same topic :
https://medium.com/coding-blocks/google-lens-firebase-54d34d7e1505

https://medium.com/coding-blocks/creating-a-credit-card-scanner-using-firebase-mlkit-5345140f6a5c

https://medium.com/coding-blocks/creating-a-qr-code-reader-using-firebase-mlkit-60bb882f95f9

https://medium.com/coding-blocks/identifying-places-in-a-provided-image-using-firebase-mlkit-fe3c918756da

https://heartbeat.fritz.ai/building-pok%C3%A9dex-in-android-using-tensorflow-lite-and-firebase-cc780848395

https://heartbeat.fritz.ai/embracing-machine-learning-as-a-mobile-developer-4ebcda58d4ac

Github repo for the code covered in the blogposts :
https://github.com/the-dagger/MLKitAndroid

Requirements

Basic Android development / knowledge of basic Kotlin.

Speaker bio

Android Developer and an avid tech blogger, Harshit is passionate about anything and everything related to Android.
He is one of the first Google certified Android developers in India and being an Open Source enthusiast, he’s also a part of various programs like Google Summer of Code and Google Code In as a Mentor.

Recently he has been inclined towards his new found love that combines his knowledge of Mobile Development with Machine Learning to create smart mobile apps.
Harshit is working with Udacity and Coding Blocks, a startup in New Delhi focusing on creating more employable talent.

Links

Slides

https://docs.google.com/presentation/d/1ibjMFCZkqbdN6Arxbt8HxKLOOAK2RahIDCyOMCuBJQI/edit?usp=sharing

Comments

Login with Twitter or Google to leave a comment