Google is bringing machine learning to Android with a newly added Neural Networks API. The NNAPI is part of the latest developer preview of Android 8.1, Oreo.
According to the company, the new API will enable developers to run computationally intensive machine learning operations on mobile devices. In addition, it provides a base layer for machine learning frameworks such as TensorFlow Lite and Caffe2, and supports inferencing such as classifying images, predicting user behavior, and selecting appropriate responses to queries.
“NNAPI is meant to be called by machine learning libraries, frameworks, and tools that let developers train their models off-device and deploy them on Android devices. Apps typically would not use NNAPI directly, but would instead directly use higher-level machine learning frameworks. These frameworks in turn could use NNAPI to perform hardware-accelerated inference operations on supported devices,” according to the API’s website.
Google says there are some trade-offs developer should be aware of: system utilizing and application size.
Other features of the Android 8.1 Developer Preview include: Android Go memory optimizations and targeting, autofill enhancements, and shared memory API.
In addition, the company announced the release of Android Studio 3.0. The latest release focuses on accelerating app development. Google recommends developers building with Android 8.1 update to Android Studio 3.0.
Android Studio 3.0 focuses on three main areas: performance issues, the Kotlin programming language, and the latest Android Oren APIs.
Google first announced support for the Kotlin programming language at its annual Google I/O conference this year. The latest update to Android Studio comes with the language support built inside the IDE. Developers can take advantage of features such as code completion and syntax highlighting.