What is Android Neural Network API?

Preview of interesting feature, that will be implemented in Android 8.1
27 October 2017   2486

Yesterday, Google started to seed to developers a new developer beta version (8.1) of Android Oreo.

The most interesting thing in 8.1 version is Neural Networks API. As Google assure, the Android Neural Networks API (NNAPI) is an Android C API designed for running computationally intensive operations for machine learning on mobile devices. NNAPI is designed to provide a base layer of functionality for higher-level machine learning frameworks (such as TensorFlow Lite, Caffe2, or others) that build and train neural networks. The API is available on all devices running Android 8.1 (API level 27) or higher.

On-device inferencing has many benefits:

  • Latency: You don’t need to send a request over a network connection and wait for a response. This can be critical for video applications that process successive frames coming from a camera.
  • Availability: The application runs even when outside of network coverage.
  • Speed: New hardware specific to neural networks processing provide significantly faster computation than with general-use CPU alone.
  • Privacy: The data does not leave the device.
  • Cost: No server farm is needed when all the computations are performed on the device.

Android Neural Network API architecture
Android Neural Network API architecture

There are also trade-offs that a developer should keep in mind:

  • System utilization: Evaluating neural networks involve a lot of computation, which could increase battery power usage. You should consider monitoring the battery health if this is a concern for your app, especially for long-running computations.
  • Application size: Pay attention to the size of your models. Models may take up multiple megabytes of space. If bundling large models in your APK would unduly impact your users, you may want to consider downloading the models after app installation, using smaller models, or running your computations in the cloud. NNAPI does not provide functionality for running models in the cloud.

This features can be very useful in situation like picture classify.

You can learn more at this page.

New Vulnerability to be Found in Google+

Due to this vulnerability it was possible to obtain private information of 52.5 million accounts
11 December 2018   108

Google decided to close the social network Google+ not in August 2019, but in April. The reason was another vulnerability in the API, due to which it was possible to obtain private information of 52.5 million accounts. The company plans to close the social network API until mid-March 2019.

By December 10, 2018, the following error information was published:

  • Third-party applications requesting access to profile data, because of the bug in the API, received permission to view information, even if it is hidden by privacy settings;
  • the names of users, their email addresses, information about occupation, age and other confidential information were at risk;
  • passwords, financial data and national identification numbers have not been compromised;
  • the company has no evidence that anyone has exploited the vulnerability;
  • the error was fixed within 6 days: from November 7 to November 13, 2018.
  • Google said it sends notifications to all users affected by the bug.

The previous data leak of Google+ users occurred in October 2018. Then about 500 thousand accounts were compromised. The attackers could get the names, email addresses, age, gender and occupation of users.