waifu2x-caffe VS ESRGAN

Windows version of the tool

I'd also recommend https://github.com/lltcggie/waifu2x-caffe/releases for a free upscaler.

Waifu2x in particular, is one of the oldest and still one of the best AI upscalers for anime and drawings in general, and it has always heavily favored Nvidia. The most known port was not the original Waifu2x by nagadomi, but Waifu2x-caffe, which was much more feature rich, and supported CUDA and cuDNN, both Nvidia exclusive. There was an OpenCL alternative which worked on AMD, but it was only as fast as Waifu2x-caffe under CUDA, but with cuDNN it still blew the OpenCL out of the water.

If u want a more detailed program you can look up their github and download it to upscale up to 10x it runs on both cpu and gpu cores so itl work on whatever pc you got. Just scroll down a bit. git hub

It's open source software just as many other projects on github are. Can't really help you translate, but the program is in many languages including English the author is just Japanese https://github.com/lltcggie/waifu2x-caffe If you scroll down you'll see some English text.

I used waifu2x, its client version https://github.com/lltcggie/waifu2x-caffe/releases. It is intended for drawn images, not realistic styles, but maybe you can find a use for this too since you use upscale in your workflow.

The settings I used for the original upscale (using waifu2x-caffe):

This seems to be based on ESRGAN which is supposed to be higher quality than waifu2x.

I downloaded https://github.com/xinntao/ESRGAN , I run it using command line, no gui.

Nerd info: Uses 3 different algorithms together: https://github.com/CompVis/stable-diffusion For generating source https://github.com/TencentARC/GFPGAN For repairing faces, mouths, eyes https://github.com/xinntao/ESRGAN For upscaling Paint.Net for manual retouching, cleanup and adjustments.

Code for https://arxiv.org/abs/1809.00219 found: https://github.com/xinntao/ESRGAN

Even if it's not a simple black and white image, you can get damn impressive results when upscaling images with GAN-based algorithms, like ESRGAN, together with a suitable pretrained model. Both are free of charge. The software can be tricky to use compared to Photoshop, though.

I use ESRGAN: https://github.com/xinntao/ESRGAN

There are two Generator architectures that are in the code, however these are actually just sitting there unused, likely from work that did not pan out because if you look at their inference code it's not used at all. Instead they directly import the plain old vanilla RRDB architecture from BasicSR yet again another Xinntao repository. Seeing the theme here.