Google open source artificial take pictures of portraiture

Owned phone pixel is one of the best cameras among flagship smartphones thanks to the techniques of artificial developed by Google in it so that can take pictures similar to what professional cameras or phones that have two cameras, but through the one camera, and soon will enable application developers making apps camera offering similar results.

Use the Google techniques of machine learning and intelligence made my use of the lens of the smartphone camera small take pictures of porter with the background of subliminal and intense focus on near objects.

I decided to Google to open source their technology under the name of DeepLab-v3+ to help developers making apps give the same effects. Of course the software made by Google open source is not identical to what I did in the phones pixels, but it helps greatly to get similar results.

Can be considered DeepLab-v3+ as a tool for chopping images using neural networks to determine the image data and the existing things of people and then work on the division of the data separation existing in the background from those in the article in order to get the effect of the bokeh effects

Of course such techniques are used by many companies, but Google specifically has developed its expertise substantially over the past five years so that it reached the level of precision can not distinguish the results of the modified software on the results of the modified gear, according to her.

Blog Google open source artificial take pictures of portraiture appeared first on the tech world.

Leave a Reply

Your email address will not be published. Required fields are marked *