Google Pushes Machine Learning Forward With Open-Source Code, Simple APIs

Google’s Android division is lowering the intimidation factor in machine learning with its solutions geared at getting developers started in the cutting-edge technology.

Head of Platform Development Relations at Google Timothy Jordan said the key to understanding machine learning is in mastering application program interfaces (APIs), which include routines, protocols and toolkits for building software applications.

’As long as you know how to use an API, you can get started using machine learning today in your apps,’ Jordan said.

Google has recently shared many of its own API models to help users get started in machine learning, including its Mobile Vision API.

’You can run it locally on a phone, you don’t even have to communicate with a server to do machine learning real time on your app. It includes APIs to detect faces, scan barcodes, [and] recognize text, all within your mobile app,’ Jordan said of the API.

Developers can find ready-to-use machine-learning models for speech, vision and translation in the Google Cloud, and the company has cloud machine-learning models in Alpha mode.

In fall 2015, Google open-sourced its artificial intelligence and machine-learning software library, which it calls TenserFlow. Google has shared the code that underwrites its deep learning models with the entire world.