"Masterpiece" button (wondering who owns the copyright for the "photos" from this preview?)
Machines have learned to generate faces that do not exist in the real world. It sometimes takes me a few seconds to understand that in front of me is not a real photograph, but an artificially created face. Sometimes I don’t understand at all that there is an artificially created photograph in front of my eyes. Actually, such an experiment was the starting point for this small note.
To understand a little about what is at stake, I advise you to read the different materials at the following links:
- https://www.izakayasushilounge.com https://thispersondoesnotexist.com/ each time the main page is updated, a new unique person’s face (portrait) is created. The image size is already 1024 * 1024, which is oh so good
- Nvidia taught the car how to create faces using given input parameters. You can watch a very interesting video clip here. here
- There are a lot of similar things. For example, the removal of excess from photographs (something that often has to be done at work in Photoshop) can be observed here
- Noise handling, link (rate moment with koala and removal of inscriptions)
All this is very interesting.
Personally, as a practicing photographer, most of my time is spent on:
- viewing all footage from the session and choosing the best shots. The task at first glance is simple, but carefully reviewing at least 1000 pictures, deciding which picture to keep from a series of pictures, and which one to delete is the most difficult task. It’s not difficult to remove, it’s difficult to take. The most difficult thing is to select photos for further detailed revision
- it takes a lot of thoughtful processing of the selected images. Everyone wants to be the most beautiful or to see their loved ones in the best possible light. All this requires a colossal investment in a 'Photoshop workbench'. It is difficult for amateur photographers to estimate the level of loading at times when several thousand frames are recruited per week. It’s not difficult to remove, it’s difficult to process
Most likely, in the future, such developments will significantly simplify and help me and other photographers in selecting the best frames and finalizing them. And also this thing should help:
- time goes on, processing skills are growing, and there is not enough time for everything. At the beginning of my work, I have a lot of interesting photos, which now, with new processing skills, would play with other colors. It would be great if I could feed the whole network of my old photos to the neural networks, and get 5-6 excellent frames at the output, which are not ashamed to show anywhere
- skinton issue will be resolved. Skinton will be anything you like, it will not depend on the sensor of the photoapp, nor on the processor, but only on the wishes of the photographer
- you can come up with many more useful things, you can voice ideas in the comments
It would be very nice if the software package quickly and efficiently, and most importantly batch processed the necessary photos with an individual approach to each individual photo. Most likely in the future such a mechanism will already be built into the cameras. Most likely there will be a complex hardware-software complex that can immediately issue photos in a certain way, with given wishes for a future photo. Again, most likely it will be possible to easily simulate the pictures of any (any!) Photographer, in any style with any effects.
Old and new will merge and the masterpiece button will still be created. This whole topic with neural networks and all the like that revolves around AI at first seems like a funny toy, but the attitude towards it changes over time. First of all, this will force photography to a new level. In the comments, as expected, there will be thoughts that no one can replace a good photographer, but, alas, in many photo tasks it will not be so.
No results
Material prepared Arkady Shapoval. Training/Consultations | Youtube | Facebook | Instagram | Twitter | Telegram
It will be the same as with any other invention of mankind. Partially will go to the needs of extermination of their own kind. Partly - to create masterpieces (this is without irony). And mainly - for fun and entertainment for millions of hamsters who need everything new and improved. Soon we will see new and improved cats, new and improved hearts and new and improved selfies from the category “me and another camel against the background of the pyramids”.
That's right!
In the future, photographs and videos can be taken immediately in the head and processed there, doing any editing with the power of thought. The organs of people will be replaceable, in fact, people will become half-borgs, so there will be no problems with vision (to fix the event). Forwarding photos, displaying images on the screen or telepathically printing.
I have such a fantasy played out when I look into the lens of a bong :)))
Soft lenses on the open and there is a retouch, for a long time already opticians have thought of everything, it remains to use it correctly.