Computational photography is one of the main strengths of Google today. The big G showed that it does not take more than one camera to have spectacular results , and that a portrait mode without the clipping problems that we see in other signatures is more than possible. Some time ago,, and now, Google has wanted to go further.
Currently, in science, it is very relevant that there is a constant large number of publications. These journal articles, books, dissertations and others, bring knowledge and evidence to the world, and allow us to progress little by little. Google has decided to publish a scientific article on the portrait mode of Google Pixel , something that we are not used to seeing, and that allows us to realize the importance of their algorithms today.
The portrait mode of the Pixel is science
If you are a member of the scientific community – this includes being a university student – you can have access to full text at, in which explain that it is possible to have portrait mode with only one camera , something in which they were pioneers, and to which the rest of the manufacturers are adding themselves – in the selfie camera.
We recommend you |
Google explains how your targeting network works , capable of separating people and objects from the background through computer photography. Google trains its AI to be able to recognize objects, people, scenes and fine details, so the portrait mode is adapted to all possible situations by default, thus avoiding the errors in clipping that most high-end make in fine details such as hair, ears, nose, glass … Instead of using a second sensor to capture information, the Pixel is able to make a semantic segmentation – thanks to certain labels that the AI recognizes – of object and background, which subsequently ends joining in a final image.
This image with portrait mode, also, uses Pixel’s automatic HDR + technology , thanks to Intel’s Visual Core chip with 2 GB of dedicated RAM, which allows to do this type of image processing. In this way, Google has wanted to contribute knowledge to science by publishing this article, which demonstrates the intentions and the pride of Google with this technology .