Blog

If you at all follow the economics of mobile platforms, then you’ve almost inevitably heard about (and likely saw) Epic’s recent viral campaign in response to Apple’s decision to ban the incredibly popular, free-to-play game Fortnite from the App Store.

You can read much more about what’s happening between Epic and Apple, but it boils down to this: The App Store has banned Fortnite for violations of its terms of service, specifically because Epic has been using its own in-game payment system. This makes some sense, given Fortnite’s status as an elite freemium gaming experience.

Continue reading “Monetizing Mobile Machine Learning”

In a previous article, I described how to utilize speech recognition in Android. This consisted of capturing user speech, processing it, and implementing it.

But what if that process is reversed, and you want to take a text input and output speech? This type of system is, so-called, text-to-speech (TTS). TTS software, in general, creates a computer-generated voice that’s also considered as an assistive technology tool.

Continue reading Make your Android apps talk with Text-To-Speech

While flashy deep learning research grabs headlines, what happens to models after they are trained is equally important. To build a great product, you need to plan for the entire lifecycle of machine learning models, from data collection and training to deployment and monitoring.

This becomes even more critical when deploying ML models outside of the cloud, directly in mobile apps where you face the unique challenges of supporting multiple platforms, hundreds of chipsets, and billions of installs.

Continue reading “The Lifecycle of Mobile Machine Learning Models”

Machine learning, deep learning, neural networks, artificial intelligence. You can’t work a day in tech without coming across one of these buzzwords.

For a developer just looking to get started, it’s hard to wade through the jargon and ever-changing tools. This post is the first in a broader series we’re working on at Heartbeat on to help people get up to speed or stay current.

Continue reading “Mobile Machine Learning 101: Glossary”

For developers, the ability to run pre-trained models on mobile signifies an important shift towards edge computing. By being able to perform data processing straight from the user’s phone, private data remains in their hands, apps run more smoothly without having to wait for cellular networks, and your company’s cloud bill is significantly reduced.

Fast, responsive apps can now run complex machine learning models. This technological shift will usher in a new wave of app development by empowering product owners and engineers to think outside the box.

Continue reading “Intro to Machine Learning on Android — How to convert a custom model to TensorFlow Lite”

In this tutorial, we’ll build a Recurrent Neural Network (RNN) in PyTorch that will classify people’s names by their languages. We assume that the reader has a basic understanding of PyTorch and machine learning in Python.

Continue reading “Detecting the Language of a Person’s Name using a PyTorch RNN”

Lens Studio can gives one the ability to create some truly amazing augmented reality effects for Snapchat. I’ve found the performance and speed with which these lenses display to be very smooth, and this is due in part to the strict requirements that Snapchat has for the quality of submitted lenses.

However, this performance comes with a price. One of the most challenging things any lens creator faces is fitting everything into the 4mb of space that is allowed.

Continue reading “Tips and Tricks to Keep your Lens Within Lens Studio’s 4mb Limit”

This is the second part in a two-part series. I suggest you read the first part first. Interact with the app via the App link

The goal of the second part of this tutorial series is to deploy the machine learning wine quality prediction web application built in part 1—this time in the cloud.

Continue reading “Deploy a Machine Learning Model as a Web Application (Part 2)”

The joy derived from having artificial intelligence models in production can’t be quantified, especially when users start interacting with the technology. AI products are the end result of iterative processes throughout a larger data science lifecycle.

Continue reading “Deploy a Machine Learning Model as a Web Application (Part 1)”