Google Research explains the Pixel 2 portrait mode
(Digital Photography Review)

Standard image (left) and depth mask (right), Image: Google Background-blurring portrait or bokeh modes have pretty much become a standard feature on dual-camera equipped phones. Similar effects can be achieved with single-lens devices but operation tends to be more cumbersome, with more manual interference required, and results less realistic than on dual-camera setups. On the the new Pixel 2 models Google has been able to implement a portrait mode on a single-lens phone that can compete with the dual-camera competition in terms of both operation and image results. Marc Levoy and Yael Pritch, two of the engineers behind the Pixel 2 portrait mode have now explained in a comprehensive post on the Google Research Blog how this has been achieved. The Google Pixel 2 offers portrait mode on both its rear-facing and front-facing cameras and uses machine learning and neural networking to generate a foreground-background segmentation. On the rear camera the initial segmentation-based depth-map is further improved using depth information generated by the Pixel 2 image sensor`s dual-pixels and a stereo algorithm. In a final step the information from both depth maps is combined to calculate the amount of blur applied to each part of the image and generate the end result. If you are interested in a more detailed description of the process you can find it, along with a range of sample images and illustrations, on the Google Research Blog. HDR+ picture without (left) and with (right) portrait mode, Image: Google