Slashdot Mirror
Researchers Devise AI System To Reduce Noise in Photos (venturebeat.com)
Researchers from Nvidia, MIT, and Aalto University are using artificial intelligence to reduce noise in photos. The team used 50,000 images from the ImageNet dataset to train its AI system for reconstructing photos, and the system is able to remove noise from an image even though it has never seen the image without noise. VentureBeat: Named Noise2Noise, the AI system was created using deep learning and draws its intelligence from 50,000 images from the ImageNet database. Each came as a clean, high-quality image without noise but was manipulated to add randomized noise. Computer-generated images and MRI scans were also used to train Noise2Noise. Denoising or noise reduction methods have been around for a long time now, but methods that utilize deep learning are a more recent phenomenon.
I have a couple of thousand images that would benefit from noise reduction. Shooting movement in low light means high ISO or blur, so I accept the noise.
If they wanted some serious training data, the whole astrophotography field is full of people that take dozens of pictures of the same thing then sample across all of them to remove noise. That means they have plenty of randomness in their noisy images and a nice clean one for comparison.
One thing I've been slowly trying to get going as a side project is exploring the use of neural networks to process raw image data.
A lot goes into processing a raw image, there is conversion of data from various color matrices, generally some sharpening, and also noise reduction. It seems like a good neural net could possibly handle all aspects and maybe do a better job if trained well, as it might spot patterns in noise or color conversion that algorithm designers to date have not (well except for recognizing color swatches and altering processing based on that... )
I was thinking to train you could just do some very accurate high res close up images of a variety of subjects that were very carefully color corrected. Then you would take images from a wider FOV or farther away, so that you could use the high-res images to determine what a "real" output pixel should be, vs whatever the result of combining various sensor data would be to produce a result.
Seems like a lot of potential here beyond just noise reduction...
"There is more worth loving than we have strength to love." - Brian Jay Stanley
With this kind opf digital processing going on, how will we know if something has been photchopped in the future? How will we expose Deepfakes? All this processing makes this sort of thing much harder....
Keep going please, I personally am looking forward to Sandra Bullocks porn releases!
Doesn't seem like a new idea or anything you need so-called 'AI' for.
Magnify and enhance sector A5.
Once again, life imitates science fiction.
I'm curious how the results stack up against commercial options like in Lightroom or Aperture. If these can reduce noise without softening the image, I'd be very interested in getting it.
I wish they'd quit with the AI and Artificial Intelligence monikers being applied to everything in tech these days. The day one of these AI's tells me that, no, it won't brew my coffee this morning because it is taking the day off is the day I might buy in to this nonsense.
It would be interesting to see a visual diff between the denoised result and the source image before the random noise was added, in order to see what kinds of artifacts were generated during the denoising process. For example, did it add any leaves to the image of the koala?
Any sufficiently unpopular but cohesive argument is indistinguishable from trolling.
It's only pseudorandom. "AI" is good at spotting patterns, even pseudorandom ones. Doesn't fare well in the real world. Experiment failed.
https://www.youtube.com/watch?...
#DeleteFacebook
They should be working on removing clothing in photos instead.
I would like to use this to denoise the prices for the SP500 and become a bazillionaire.
For a photographer, being able to push your ISO up crazy, and then be able to fix later in Photoshop or whatever program adapts this technology would be amazing for low light situations.
Do you think this will work on upskirt photos? Asking for a big, wet, orange friend.
You are welcome on my lawn.
While the images they have shown as examples are really pretty impressive, given that they're using a training set of Image A versus Image A Plus Noise, the problem is akin to blind source separation (BSS). There's been quite a lot of work done on BSS, much of which is very impressive (and based on neural nets).
The critical issue is to see what happens when they take a real photograph that has not been adulterated to add noise, and improve that. Will their model of a noiseless source image with additive noise still hold? The article doesn't touch upon that critical test, unfortunately.
The results they show are very, very cool, though. And if they hold up for MRI work, it would be a game-changer in the medical field. The article shows an MRI adulterated with noise, their recovered image, and the noiseless ground truth. A better test would be to take an MRI that was scanned for too short a time (and thus is noisy), and compare their extraction against an MRI with identical scanning parameters, except for normal imaging time. MRI magnet time is expensive; if it can be reduced by 50% and get equivalent image quality, that's a huge advance.
Put my fist through my alarm clock with its ding-dong death inside my ear. - The Blackjacks.
I have a few Japanese videos I'd like to process with this!
Here's a program (in development since at least three years ago) which uses neural networks to upscale and de-noise anime-style art: https://github.com/nagadomi/wa...
In other words, you think heuristics could do a better job than software that understands the specific properties of the underlying sensor.
You don't have to have just one approach, and could combine both.
However, I don't think you are understand what I am describing. The training data would come from the same sensor, I'm not talking about arbitrary images here - so the "heuristics" would in fact be learning based on the properties of the sensor for the raw data it would be working with. In fact it probably would be better to have several or a dozen different cameras using the same sensor producing higher res training data to avoid training too much against one particular sensor vs. images from multiple sensors of the same type...
The real potential problem in fact is not that it does not "understand the properties" of a sensor, it's that the approach may not be easily generalized outside a specific sensor, possibly even a specific model of camera (each kind of camera has different kinds of electronic noise mixed with sensor data recorded).
It puzzles me why you are so dismissive of the idea to the point of crass insult, when image processing and even generation based around neural networks has worked astoundingly well to date.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
Photoshop can also reduce noice from a picture it have never seen before.
No, they have developed an *algorithm* to reduce noise. Photoshop and similar have used algorithms for this since the dawn of time. Automating it does not make it a different beast, and thus far, Adobe's own 'AI' experiments are pretty much just filters that may or may not offer results as good as the old ones in trained hands. Yawn. Serious, serious yawn. And anyone expecting something like this to magically make your photos look like they weren't shot on a digital camera (noise is endemic to digital photography), good luck.
Nowadays a good pro-camera sensor shows almost no visible noise in rather dark conditions (using high ISO to keep taking photos under 20 ms). Not perfect yet, but ISO improved by a factor of 10 in 10 years. Negative had their time, and some important professional features did not evolve in pro digital camera thanks to pro-photographers unable to catch up with progress ; they'll keep talking about noise in 10 years when nobody cares anymore.
Slashdot, fix the reply notifications... You won't get away with it...
https://groups.csail.mit.edu/graphics/demosaicnet/data/demosaic.pdf
Using Deep Learning to de-noise images is cool, but not new.
3D ray tracing shoots out photons with a certain degree of randomness to build up the image.
The more photons you collect, the less grainy the picture gets, which is great for this type of training because you can generate the training data to as high or low a quality as you like.
The end result is a black box you run your image through that maybe cuts your render time in half (I don't remember what the actual improvement rates are), essentially for free.
... noise-cancelling headphones while viewing food photos on Facebook.
It little behooves the best of us to comment on the rest of us.
Improved ISO doesn't do anything to reduce shot noise. No matter how perfectly the camera counts photons, the shot noise will never get any lower.
In low light conditions, most of the noise on a good camera is already shot noise.
This is really cool, it would lead to clearer pictures of high quality Download Antman and the wasp Movie here >>> https://www.naijadailyfeed.com...
I love to blog http://www.naijadailyfeed.com
It will not create data out of thin air. The best it can do is guess the contents of a pixel based on its surroundings. This is useful, to be sure, but should never be used for forensic evidence such as recreating a suspect's face.
People forget that a few years ago Google unveiled a AI model that would reconstruct human faces based on an 8x8 blurred image. While the reconstructed faces looked impressive, it turns out that they bore little basis in reality. Black people were reconstructed as white because... you know... the training data contained more white people.
AI is visually impressive. But that's all it is. Cosmetic.
How about not limiting your thinking to your field of interest. Would this not be useful to improve X-Ray scans while keeping X-Ray doses low?
I ran the audio version of this noise-reducing software on Lou Reed's "Metal Machine Music" and ended up with a telephone dial tone.
You are in a maze of twisty little passages, all alike.
The Jet Propulsion Laboratory process image from spacecraft to space from many planetary probes. It will be extremely cool they provide them a site license to use this tech for NASA!