Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Neural Enhance – Super Resolution for images using deep learning

github.com

Post reply on HN
Print view View original

21–30 of 53 posts

Re: Neural Enhance – Super Resolution for images using deep learning

#21

Post by andai »

ingenter wrote: post #20

Here's a list of various image interpolation techniques, with similar goal: - http://www.wisdom.weizmann.ac.il/~vision/SingleImageSR.html - http://chiranjivi.tripod.com/EDITut.html - http://www.tecnick.com/pagefiles/appunti/iNEDI_tesi_Nicola_A... - http://www.eurasip.org/Proceedings/Eusipco/Eusipco2009/conte... - http://bengal.missouri.edu/~kes25c/ http://bengal.missouri.edu/~kes25c/nnedi3.zip http://forum.doom9.org/…

My question is offtopic, but how do you keep lists of urls like that? Do you just use text files? I'm struggling with too much to read

Re: Neural Enhance – Super Resolution for images using deep learning

#24

Post by Keyframe »

Leynos wrote: post #13

A friend of mine suggested that an approach similar to this could be used to upscale old standard definition TV shows (specifically, those shot on video rather than film). I'd imagine that multiple specially trained networks would be employed for different parts of the image (trained on pictures of individual performers or types of set/background). Pleased to see that this is possible. Is there anyone doing something…

Is there anyone doing something along those lines already?

I have a side business doing Film restoration and am not aware of any solution like that. Probably the best upscale solution there is is from Teranex, acquired by BlackMagic Design. Evertz probably also has something in their offering.

Re: Neural Enhance – Super Resolution for images using deep learning

#25

Post by kuschku »

andai wrote: post #23
kuschku wrote: post #6

How much RAM do I need to run this? It dies for me with a MemoryAllocation Error after eating 28GB of RAM…

Wow how much do you have?

On this system I’ve got 32GB, of which about 2GB were used by the OS itself, and another 2GB by firefox, that’s why it stopped at around 28GB.

Re: Neural Enhance – Super Resolution for images using deep learning

#26

Post by semi-extrinsic »

placebo wrote: post #15

It definitely makes a significant qualitative improvement, making the picture appear more in sync with what our brain interprets as a higher resolution picture, but my first thought is whether this particular example goes beyond aesthetics. Is there really any instance where this method could for instance turn an unintelligible picture of a license plate to something in which the characters can be recognised? More ge…

I imagine that if you trained this on a set of license plate photos, it would be able to enhance license plates illegible to an untrained human such that they're readable. However, I doubt it would be better than a human specifically trained at this task.

I've seen some videos from Cold War satellite photo analysts, and the way they can look at some tiny gray blobs and go "That's a T-64 tank, that's a T-62 tank, that's an SA2 launcher" etc.

Re: Neural Enhance – Super Resolution for images using deep learning

#27

Post by WhitneyLand »

This approach can be equaled or bettered with no machine learning.

This example allows easy comparison between common techniques. Choose image 7 to see an example with a person: https://dl.dropboxusercontent.com/u/2810224/Homepage/publica...

Re: Neural Enhance – Super Resolution for images using deep learning

#28

Post by tgb »

WhitneyLand wrote: post #27

This approach can be equaled or bettered with no machine learning. This example allows easy comparison between common techniques. Choose image 7 to see an example with a person: https://dl.dropboxusercontent.com/u/2810224/Homepage/publica...

While those results are nice too, they really don't seem comparable to the ones given here. The look is too different.

Re: Neural Enhance – Super Resolution for images using deep learning

#29

Post by Houshalter »

Leynos wrote: post #13

A friend of mine suggested that an approach similar to this could be used to upscale old standard definition TV shows (specifically, those shot on video rather than film). I'd imagine that multiple specially trained networks would be employed for different parts of the image (trained on pictures of individual performers or types of set/background). Pleased to see that this is possible. Is there anyone doing something…

It should work, I don't think you need to bother with training it on individual performers. Someone made a thing like this to improve low res anime, that worked well.

In theory you could use this to increase temporal resolution as well. Turn 24 fps movies into 60 fps, and upscale regular HD to 4k.

Re: Neural Enhance – Super Resolution for images using deep learning

#30

Post by kartan »

This technique is akin to hire an artists to draw a high resolution version of your pixelated photos.

A good example of this is "They Of The Tentacle Remastered" (http://dott.doublefine.com/). The new game looks extremely similar to the old one but it has been redrawn.

As some one suggested you should be able to take an old TV show, train the neural network with HD pictures of the cast. And let it redraw it in its own "artistic" interpretation of the images.

Post reply on HN