Overclock.net banner

1 - 6 of 6 Posts

·
Banned
Joined
·
15,763 Posts
Discussion Starter #1
Imaging in low light is challenging due to low photon count and low SNR. Short-exposure images suffer from noise, while long exposure can induce blur and is often impractical. A variety of denoising, deblurring, and enhancement techniques have been proposed, but their effectiveness is limited in extreme conditions, such as video-rate imaging at night.
This is mind blowing.

Source

Source 2


 

·
Performance is the bible
Joined
·
7,134 Posts
As editing photos, this is in the simplest terms, just a better noise reduction tool that exist in almost every photo editing software, with maybe better algorithms and more power than your average editing desktop.
I do note that in some of the examples, the algorithm works very nicely, but it adds a lot of distortions and washing of colour, which is similar to many other noise reduction tools.
But overall it is a very nice progress, though I think some of their examples are a bit off or a bit misleading, since they can't just take a black seen raw and make it visible. The raw has to have enough data for them to clean and make it visible, even though our screen can't see the full 14bit data from the raw when it extracted as is into an 8bit jpeg.
I also don't see them comparing their algorithms and noise reductions compared to other tools that exist in the market, but mostly comparisons of jpegs out of camera.
 

·
Frog Blast The Vent Core
Joined
·
6,114 Posts
As editing photos, this is in the simplest terms, just a better noise reduction tool that exist in almost every photo editing software, with maybe better algorithms and more power than your average editing desktop.
I do note that in some of the examples, the algorithm works very nicely, but it adds a lot of distortions and washing of colour, which is similar to many other noise reduction tools.
But overall it is a very nice progress, though I think some of their examples are a bit off or a bit misleading, since they can't just take a black seen raw and make it visible. The raw has to have enough data for them to clean and make it visible, even though our screen can't see the full 14bit data from the raw when it extracted as is into an 8bit jpeg.
I also don't see them comparing their algorithms and noise reductions compared to other tools that exist in the market, but mostly comparisons of jpegs out of camera.

That's like saying that your smartphone is just a better communication tool that has existed in almost every household for the last several decades.

It's not as misleading as you claim, as the standard gain and noise reduction features you mention as being common end up making the pictures look completely awful. Yes, it's not perfect, but it is DRAMATICALLY better.

I have some experience with image processing and analysis in a professional setting, and this work is truly impressive. I do wonder whether they'd be able to get the processing fast enough for video. Would be a total gamechanger for night vision technology.

I must say I find this sort of armchair sniping really irritating. This is phenomenal work, and just because you don't appreciate the effort does not make it not impressive.
 

·
Banned
Joined
·
15,763 Posts
Discussion Starter #5
That's like saying that your smartphone is just a better communication tool that has existed in almost every household for the last several decades.

It's not as misleading as you claim, as the standard gain and noise reduction features you mention as being common end up making the pictures look completely awful. Yes, it's not perfect, but it is DRAMATICALLY better.

I have some experience with image processing and analysis in a professional setting, and this work is truly impressive. I do wonder whether they'd be able to get the processing fast enough for video. Would be a total gamechanger for night vision technology.

I must say I find this sort of armchair sniping really irritating. This is phenomenal work, and just because you don't appreciate the effort does not make it not impressive.
I only wish I was this smart as these nerds that did it.Most people don't appreciate the work put into it. Often inventors are the most discredited. To me this technology is a huge breakthrough in taking photos in the night where lighting won't be a problem anymore. A simple algorithm will fix all that noise and other artifacts on the screen.
 

·
Registered
Joined
·
2,797 Posts
This is wonderful. Can't wait to see the new smartphones implementing this.
 
1 - 6 of 6 Posts
Top