The great thing about pictures is that they capture memories that can last a lifetime. It’s always great to look back on old family photos and see how far technology has come from the black and white, low-resolution images.
Now, we’re getting more information on a new free tool that can actually restore old images by producing slightly newer versions.
While it does do the job impeccably, there is always this small risk of getting a picture at hand with a slight change in the actual identity.
Thanks to the power of AI-powered technology, not only can you get new pictures, but restore old ones as well. The new free-of-cost tool is brought to us by the efforts of researchers hailing from Tencent.
The technology can merge data from two separate models of AI and combine to fill up the missing elements of an image using realistic detail. Within a few seconds, low-quality images are given the kick they need to appear as good as new. And the same goes for pictures that were damaged.
The way accuracy is maintained with quality enhancement is just too good to be true for some experts that call the ordeal nothing less than amazement.
Old and conventional techniques that were in use today have the capability to get your images back to looking new by determining the difference between both artificial with what’s real. And in the end, you’re left with nothing but poor quality images, as revealed by scientists researching such procedures.
This new technique is designed to make use of pre-trained visions of models that give a heads-up of what’s happening throughout the process of image production. The whole idea is to preserve the true identity of those individuals seen in the picture.
There is a special emphasis drawn upon specific facial features such as the mouth, nose, or even the eyes.
GFP-GAN has even allocated a demo that interested users can try out to see the master effects in seconds. This comes alongside the posting of a code that enables anyone to make use of the revolutionary technology in their design projects too.
It should be remembered that despite the huge benefits, the project is still limited by the current levels of artificial technology available. And while we do love how accurate it can be, it really ends up putting forward educated guesses on what’s missing and what’s not.
Therefore, this is the main reason why researchers are issuing warnings to users that there could be a possibility of small changes in identity and even a resolution that is less than what some may expect.
In case you have plans to get such images of your ancestors printed to fit across a poster, well, don’t because there could be some differences.
Whatever the case may be, we’re sold. Everything over here that we’ve discussed regarding the tool appears to be promising. We’re hinting toward a future where images can be salvaged instead of memories being erased forever.
H/T: PP
Read next: AI-Based Art Is The New Trend, But The Internet Questions Whether Or Not It Is Harmful
Now, we’re getting more information on a new free tool that can actually restore old images by producing slightly newer versions.
While it does do the job impeccably, there is always this small risk of getting a picture at hand with a slight change in the actual identity.
Thanks to the power of AI-powered technology, not only can you get new pictures, but restore old ones as well. The new free-of-cost tool is brought to us by the efforts of researchers hailing from Tencent.
The technology can merge data from two separate models of AI and combine to fill up the missing elements of an image using realistic detail. Within a few seconds, low-quality images are given the kick they need to appear as good as new. And the same goes for pictures that were damaged.
The way accuracy is maintained with quality enhancement is just too good to be true for some experts that call the ordeal nothing less than amazement.
Old and conventional techniques that were in use today have the capability to get your images back to looking new by determining the difference between both artificial with what’s real. And in the end, you’re left with nothing but poor quality images, as revealed by scientists researching such procedures.
This new technique is designed to make use of pre-trained visions of models that give a heads-up of what’s happening throughout the process of image production. The whole idea is to preserve the true identity of those individuals seen in the picture.
There is a special emphasis drawn upon specific facial features such as the mouth, nose, or even the eyes.
GFP-GAN has even allocated a demo that interested users can try out to see the master effects in seconds. This comes alongside the posting of a code that enables anyone to make use of the revolutionary technology in their design projects too.
It should be remembered that despite the huge benefits, the project is still limited by the current levels of artificial technology available. And while we do love how accurate it can be, it really ends up putting forward educated guesses on what’s missing and what’s not.
Therefore, this is the main reason why researchers are issuing warnings to users that there could be a possibility of small changes in identity and even a resolution that is less than what some may expect.
In case you have plans to get such images of your ancestors printed to fit across a poster, well, don’t because there could be some differences.
Whatever the case may be, we’re sold. Everything over here that we’ve discussed regarding the tool appears to be promising. We’re hinting toward a future where images can be salvaged instead of memories being erased forever.
H/T: PP
Read next: AI-Based Art Is The New Trend, But The Internet Questions Whether Or Not It Is Harmful
