Google has detailed its Pixel 6 and Pixel 6 Pro smartphones, which emphasise on-device AI features and are the first to use its in-house-designed Tensor processor, at an online event overnight.
Differing quite a bit from last year’s Pixel 5 in appearance, the phones have a two-colour finish with a black horizontal camera bar on the rear, an in-display fingerprint reader and a centred hole punch for the front-facing camera.
The Pixel 6 comes in grey, blue/green and pink, while the Pixel 6 Pro comes in grey, white and yellow.
The standard Pixel 6 features a 6.4-inch 90Hz OLED screen and a pair of cameras — one wide and one ultrawide — while the Pixel 6 Pro aims to match the latest specs from Apple and Samsung with a 6.7-inch 120Hz OLED and the addition of a periscope telephoto lens at 4x zoom. In Australia, the phones will start at $999 and $1299 respectively.
The phones will come with Android 12, which Google has customised with a new interface design that adapts and personalises colours and widgets used across the device. The company was also keen to point out how the combination of Android 12 and Tensor enabled on-device AI tasks not possible on other phones.
The Pixel 6’s magic eraser feature can remove people from photos and fill in the space.
One of those tasks shown off during the event was a new Google Photos feature called Magic Eraser. Users can tap people and objects in their photos, and the Pixel 6 will use AI to remove them and fill in the background to make it appear as if they were never there.
Another AI-powered task expands Pixel’s on-device translation feature to work in more contexts. Google showed how text messages can be translated as they’re sent and received, so for example an English speaker and a French speaker can have a conversation and each get messages in their native languages. It currently works for 11 languages and across apps including WhatsApp, Signal, Line and Twitter.
Loading
Live Caption, which instantly transcribes the speech in any video or audio content on the phone, can now also translate at the same time, so you can watch a Japanese livestream and get English captions generated for example. Improvements to the on-device language model also make voice typing more accurate, Google said.