A few weeks ago, Google Photos introduced cinematic photographs in Spain, a new option capable of converting our photographs by adding a three-dimensional movement effect that gives the sensation of getting closer to the main subject of the photograph.
How did Google manage to create a mode capable of converting a still image into a small moving video? The company has explained how they have succeeded.
This is how Google Photos cinematic photos work
Cinematic Photos is a new mode present in Google Photos capable of giving a new layer of depth to our images, achieving an effect that simulates as if we were getting closer with the camera.
Like every new innovation in the field of mobile photography, artificial intelligence has been used to achieve this tool, and this time there are several steps it follows. The first one is similar to that used by technologies such as augmented reality or portrait mode, and is to create a depth map. The depth map is in charge of identifying the main subject and separating it from the background.
Once the image has been segmented, the next step is to simulate the three-dimensional movement of a camera by moving the background and analyzing the rear distortions that occur in the movement.
With the information obtained from the multiple scenes, the next step is to achieve a correction of the three-dimensional effect, comparing all the samples generated from the 3D animation, with which we would achieve the zoom effect.
The last step of cinematic photos is the framing of moving photography so that we are the center of attention in this spectacular animation.
Seen from this perspective it seems a simple process, but within the development they have implemented several Machine Learning technologies. If you want to know how it works in greater depth and you understand a lot of English, we recommend you read the Google AI article.
The entry This is how the spectacular cinematic photographs of Google Photos work appears first in The Free Android.
