We all know very well that almost no profession will be immune to advances in artificial intelligence. Hence, the tech giant Google has just taught a machine (AI) simply to edit photos as a professional, and the first results are really promising.
Google Teaches AI To Edit Photos As A Professional — Check Out The Results
Almost no profession will be immune to advances in artificial intelligence. The tech giant Google has just taught a machine to edit photos as a professional, and the first results are promising.
What sets a professional edition apart from an amateur? It is clear that the professional has more study and experience, but subjective factors are often taken into account when evaluating an image.
The tech giant Google trained the machine (of course, Artificial Intelligence) to learn these factors: it was presented to a collection of raw, professionally edited photo pairs, and then invited the AI simply to apply the techniques on its own.
In the fire test, artificial intelligence “edited” about 40,000 panoramic photos taken from the tech giant Google’s application Street View and taken in parks in Canada, the United States, and Italy. The computer trimmed the photos, improved saturation, dynamic range and lighting, and also defined the composition of the scene.
To better determine the quality of photo editing, the tech giant Google called professional human photographers, who gave a score of 1 to 4, with “1” being a common photo taken without concern for composition or lighting, and “4” is a photo Edited by a professional. Result: 40% of the images were judged as semi-professional or professional.
The intention, of course, is not to completely replace a professional photographer – but ordinary users can already expect the news to create better photos in Google Photos and other software in the future.
With the arrival of superintelligent machines (Artificial Intelligence), human labor will be increasingly unnecessary. Even that goes for the creative and intellectual work, too! Do not think that we are immune to change. The algorithms are designed to learn similarly to the human brain. So if we can, “they” will also – only they will do better and faster.
So, what do you think about this machine learning? Simply share your views and thoughts in the comment section below.