Google reveals Pixel's portrait mode secret; open-sources the AI tool

The AI tool

Earlier this week, the company in a blog post announced that it has open-sourced a lump of code name DeepLab-v3+.

The tool is built using convolutional neural networks: a machine learning method that excels in analyzing visual data. The tool analyzes the subjects in a picture and divides the foreground elements from the background elements.

Top 10 Tips & Tricks for Google Pixel 2 and Pixel 2 XL

What's the use?

This sounds normal, but it is actually a very useful skill for cameras, and Google makes use of the tool for its portrait mode images on the Pixel smartphones. The portrait mode captures photos with a bokeh effect that blurs the background of the subject in focus. The feature was first popularized by the iPhone, but it's worth noting that the Pixel does it with only one lens while the iPhone uses two lenses.

Google hopes that by sharing the system with the public, "other groups in academia and industry [will be able] to reproduce and further improve" the company's work.

What else?

Besides, Google is also reportedly working on developing new gesture control features for its wireless earphones Pixel Buds. According to some reports, Google appears to develop a triple-tap feature which turns off the earbuds instantly. This feature is expected to improve the battery life of the Pixel Buds.

Further, Google has also renamed its Smart Detection feature to "In-ear-detection". The Smart detection feature enables the Pixel Buds to pause the music playback as soon as the earbuds are removed from the ears. As of now the Pixel Buds allows a user to only reject and answer calls along with activating Google Assistant, and control volume audio playback using gestures.

It would be interesting to see what other feature Google has in store for us. The Google Pixel Buds were released back in October 2017 and were made available for more countries in February 2018.