nuclai

5 ms·

1.
by nuclai 10y ago
Yes, it sounds possible with this code — but would require training a new network. Do you have a link to some examples?
2.
by nuclai 10y ago
(Author here.) Maybe it's worth moving to a GitHub issue. Try `--model=small`. The demo server limits the number of pixels to around 320x200 or 256x256 and can do only 4 at the same time to fit in RAM.
3.
by nuclai 10y ago
(Author here.) Unlike most other non generative adversarial network (GAN) approaches to super-resolution, it does try to inject high-frequency detail; see the faces example on GitHub. But I tuned down that parameter in the released models a
4.
by nuclai 10y ago
(Author here.) If you have the luxury to train on domain-specific textures, the results will definitely be better. That's why I included all the training code in the repository as well—to allow for this kind of solution. If you scroll
5.
by nuclai 10y ago
(Author here.) Did you see the faces example on the GitHub page? It was a domain-specific network trained adversarially for that purpose, but I have yet to see any super-resolution of that quality with or without machine learning. Most othe
6.
by nuclai 10y ago
(Author here.) Yeah, I knew this would come up but decided to proceed with the pixelated comparison anyway. I couldn't get the GIFs to reflect the results because of 8-bit quantization/dithering. The images show the neural networ
7.
by nuclai 10y ago
(Author here.) Absolutely! Using multiple super-resolution networks, not only continuity would present problems, but also blending between different regions. I agree there's a lot of value for domain-specific networks here, as you can
8.
by nuclai 10y ago
(Author here.) My biggest insight from this project is that super-resolution with neural networks benefits significantly from being domain specific. If you train on broader datasets, it does pretty well but has to make compromises. Many rec
9.
by nuclai 11y ago
Almost nobody in deep learning uses OpenCL. All the DL frameworks primarily focus on CUDA and that's where you get the best performance. OpenCL is off the beaten path and you pay for it in every way: support, performance, reliability.
10.
by nuclai 11y ago
You can use any image as source, but to create annotations you have to do that yourself currently. Using simple segmentation libraries (or clustering) can do a good job for certain images, or look at better solutions for semantic segmentati
11.
by nuclai 11y ago
Oh, absolutely. It's an idea whose time had come ;-)
12.
by nuclai 11y ago
From that perspective, this research is two steps further than Neural Style, I wrote about it yesterday here: http://nucl.ai/blog/neural-doodles/ First, the paper I call "Neural Patches" (Li, January 201
13.
by nuclai 11y ago
The semantic map remains static during the optimization, so it can be provided as a pre-computation (e.g. pixel labeling, semantic segmentation, etc.) or done by hand. The ones in the repository are done manually, but now experimenting with
14.
by nuclai 11y ago
The algorithm does the same thing every time (it's triggered on request), only the input is changed by the human modifying the doodle—as shown in the video. The output gets better because through iteration the glitches are removed incr
15.
by nuclai 11y ago
Exactly, the doodling is done by humans and the machine paints the HD images based on Renoir's original. I've edited the blog post to clarify.
16.
by nuclai 11y ago
The research is based on work I did writing and improving @DeepForger ( http://twitter.com/deepforger ), an online service for "basic" style transfer. The GitHub is a standalone version for learning and education, w
17.
by nuclai 11y ago
> This code does a little of that. Actually, the code does none of that ;-) All of the semantics are provided by the users: either as manual annotations or by plugging in an existing architecture for semantic segmentation / pixel la
18.
by nuclai 11y ago
Thanks for clarifying, I'll update the README. The research paper does a better job of explaining this with its figures! The algorithm can only reuse combinations of patterns that it knows about, it can do extrapolation but it often en
19.
by nuclai 11y ago
You can specify two pairs of images (content+annotation) and it'll transfer the style from one to another as consistently as possible. The down side is that you need to find an algorithm, neural network, or person to create the annotat
20.
by nuclai 11y ago
No, do you have any good ones? As long as entire sections are colored (not just lines), and those colors match with the annotations of another image, it should work fine!
21.
by nuclai 11y ago
Both images have their patterns extracted by the NN, and the optimization then tries to match the best patches from one image with the other, performing gradient descent to adjust the pixel values from a random start image.
22.
by nuclai 11y ago
It's a pre-trained network on image classification dataset from 2014 called ImageNet. The network is called VGG, paper is here: http://arxiv.org/abs/1409.1556 There's no additional training apart from that. T
23.
by nuclai 11y ago
(Author here.) For details, the research paper is linked on the GitHub page: http://arxiv.org/abs/1603.01768 For a video and higher-level overview see my article from yesterday: http://nucl.ai/blog&#x2F